Masarykova univerzita v Brně
Lékařská fakulta
Studium elektrofyziologických projevů vyšších funkcí
mozku člověka pomocí intracerebrálních elektrod
Habilitační práce
MUDr. Robert Roman, Ph.D.
Brno 2018
2
Mé upřímné poděkování za odbornou spolupráci, cenné připomínky a plodné diskuze,
které doprovázely všechny publikované práce, patří zejména prof. M. Kukletovi
(in memoriam) a prof. M. Brázdilovi, za všestrannou pomoc při registracích a zpracování iEEG
dat pak spoluautorům a kolegům, se kterými jsem měl a stále mám tu čest spolupracovat –
dovolím si uvést alespoň dr. A. Damborskou, ing. J. Chládka, ing. P. Juráka a ing. P. Daniela.
Za podporu a vytvoření podmínek pro moji vědeckou práci patří poděkování také
bývalým přednostům Fyziologického ústavu LF MU v Brně prof. N. Honzíkové a prof.
B. Fišerovi (in memoriam).
Za pochopení, podporu a trpělivost děkuji své rodině.
3
Souhrn
Schopnost člověka odpovědět dohodnutým způsobem na definovaný podnět je
podmíněna řadou mozkových procesů, které jsou podkladem senzorického zpracování podnětů,
kognitivních a exekutivních funkcí. V předkládaném souboru prací jsme se zaměřili na studium
jejich elektrofyziologických korelátů registrovaných intracerebrálně u pacientů
s farmakorezistentí formou epilepsie. Tito pacienti souhlasili s účastí ve studiích, schválených
etickou komisí Masarykovy univerzity, probíhajících v rámci epileptochirurgické léčby
epilepsie na 1. neurologické klinice u Sv. Anny v Brně. V průběhu hospitalizace pacienti
prováděli různé kognitivní úkoly při současné intracerebrální registraci EEG. Výzkumně se
přímá registrace elektrické aktivity neuronálních populací lidského mozku využívá pouze
na několika specializovaných světových pracovištích. Umožňuje však vyšetřit pouze omezený
rozsah oblastí mozku, protože hluboké elektrody jsou implantovány v limitovaném počtu a
výhradně na základě klinické potřeby bez jakékoli vazby na výzkumné úkoly.
V úvodní části habilitační práce je popsána metodika intracerebrální (intrakraniální)
registrace elektrické aktivity mozku člověka a její výzkumné využití. V navazující části jsou
komentovány výsledky prací, zaměřených na studium jednotlivých složek elektrofyziologické
odpovědi mozku snímané v průběhu kognitivních úkolů. Základním přístupem byla analýza
evokovaných potenciálů, označovaných jako „event-related potentials“ (ERPs), které získáme
zprůměrněním obvykle několika desítek EEG odpovědí od okamžiku prezentace podnětů.
V nejvíce studovaném zrakovém oddball úkolu byla náhodně prezentována písmena „X“
(terčové podněty) a „O“ (neterčové podněty) v poměru 1:5. Pacienti byli instruováni reagovat
stiskem tlačítka pouze na výskyt terčového podnětu, který současně počítali. Výsledky našich
studií potvrdily, že terčové a neterčové podněty jsou v mozku zpracovávány dvěma různými
způsoby. První, obecnější způsob zahrnuje procesy senzorické, pozornostní, paměťové a
rozhodovací, které jsou součástí základní analýzy externích podnětů bez ohledu na jejich
význam. V řadě vyšetřených struktur (např. gyrus cinguli, gyrus parahipokampalis, gyrus
temporalis superior, medius a inferior, hipokampus a frontoorbitální kortex) by tomu odpovídal
nález identických ERP komponent po obou typech podnětů. Druhý způsob zpracování již
zohledňuje asociaci podnětu s dohodnutou odpovědí, případně jeho méně častý výskyt v rámci
úkolu, a zahrnuje opět kognitivní, ale také exekutivní a kontrolní funkce. Elektrofyziologickým
korelátem těchto funkcí jsou pravděpodobně jednotlivé kognitivní komponenty ERPs, např.
vlna P3. Na základě analýzy časového vztahu této vlny k okamžiku prezentace podnětu a
4
motorické odpovědi se nám podařilo prokázat, že intracerebrální vlna P3 je heterogenní
fenomén. Velmi pravděpodobně odráží buď vyhodnocení významu podnětu, nebo exekutivní
funkce anebo pozornostní a integrativní funkce. Ve studii zaměřené na analýzu kognitivního
ERP generovaného v hipokampu člověka v průběhu jednoduchého senzorimotorického úkolu
jsme následně ukázali, že hipokampální kognitivní potenciál s latencí kolem 420 ms nesouvisí
s provedením motorické odpovědi a zřejmě odráží vyhodnocení významu podnětu v kontextu
experimentální situace.
V další studii jsme se zaměřili na lokalizaci neuronálních populací, které v průběhu
oddball úkolu po prezentaci podnětů vykazují fázově nekoherentní změny EEG aktivity.
Signifikantní změny amplitudy EEG oscilací ve frekvenčním pásmu alfa a beta byly
identifikovány v různých oblastech mozku bez specifické vazby na konkrétní struktury.
Po terčových podnětech jsme častěji pozorovali nárůst výkonu EEG oscilací v alfa pásmu, který
by mohl ukazovat na aktivaci neuronálních sítí zajišťujících motorickou odpověď.
Po neterčových podnětech byl převažujícím nálezem pokles výkonu EEG oscilací v alfa i beta
pásmu, který by mohl odrážet převahu inhibičních procesů souvisejících s neprovedením stisku
tlačítka. Identifikované změny výkonu jsou zjevně vázány na odpověď subjektů, jejich jasná
interpretace však doposud chybí.
Intracerebrální registraci EEG dat jsme také využili k mapování potenciálů, které
odrážejí aktivaci neuronálních systémů zajišťujících kontrolu správnosti odpovědi jedince
na podněty z jeho okolí. Generátory tzv. „error“ potenciálů vázaných na chybnou odpověď
subjektu v Go/NoGo úkolu jsme našli v řadě oblastí frontálního a temporálního laloku, např.
v gyrus cinguli anterior (ACC), gyrus frontalis medialis, dorzolaterálním prefrontálním kortexu
(DLPFC), orbitofrontálním a laterálním temporálním kortexu. Z našich výsledků navíc
vyplynulo, že kaudální a rostrální část ACC patrně hrají odlišnou roli při zpracování chybné
odpovědi. V další studii jsme se zaměřili na lokalizaci jiné skupiny potenciálů vyvolaných
v oddball úkolu až po provedení odpovědi správné. Opět se ukázalo se, že v rámci vyšetřených
mozkových struktur existují minimálně dvě funkčně odlišné neuronální sítě, přičemž účast
jedné byla specifická pouze pro terčové odpovědi, u druhé byla její aktivace na typu podnětu
nezávislá. Navíc se podařilo identifikovat některá místa v ACC, DLPFC, primárním
motorickém a laterálním temporálním kortexu, která se podílela současně i na hodnocení
odpovědí chybných. Z výsledků výše uvedených prací vyplývá, že jedna anatomická struktura
může být součástí několika neuronálních sítí, které jsou strukturálním podkladem různých
funkcí.
5
Intracerebrální EEG data snímaná ve zrakovém oddball úkolu nám však také umožnila
studovat hierarchicky vyšší úroveň organizace mozkových funkcí, konkrétně funkční
konektivitu aktivovaných oblastí. Výpočtem korelací EEG signálů ve frekvenčním pásmu beta
se ukázalo, že nejvýznamnější roli při komunikaci mezi frontálními a temporálními oblastmi
mozku hraje gyrus cinguli. S použitím výpočtu „directed transformation function“ se nám
podařilo také demonstrovat, že při zpracování podnětů spojených s aktivní odpovědí se v rámci
frontálního laloku informace šíří směrem z gyrus cinguli do dorzolaterálního prefrontálního
kortexu. Výsledky těchto studií dokládají klíčové postavení předního gyrus cinguli při řízení
chování jedince.
6
Summary
The human ability to respond to specific stimuli arises from neuronal activity associated
with sensory processing of information as well as cognitive and executive functions. The
presented set of original papers studies intracerebrally recorded electrophysiological correlates
of neuronal processing in patients with medically intractable epilepsy. All studies were
approved by the Ethical Committee of Masaryk University and all patients agreed to participate
and perform several cognitive tasks while their intracerebral EEG was continuously recorded
during their stay in the hospital. Such direct investigation of the electrical activity of the human
brain is possible only at very few clinical departments in the world. The possible limitation
of the intracerebral EEG method, however, is the fact that the depth electrodes are implanted
solely on clinical grounds and without any reference to the experimental protocols.
The first part of the text introduces the method of intracranial electroencephalography
and explains its value for research. The following part includes comments of our papers where
we examined different components of the electrophysiological brain responses recorded during
various cognitive tasks. We used the analysis of evoked potentials, i.e. event-related potentials
(ERPs), obtained by averaging of several EEG epochs triggered by the presented stimuli. In the
most investigated oddball paradigm, the letter “X” was used as a target stimulus and the letter
“O” as a nontarget stimulus in the ratio 1:5. Patients were instructed to press a button after the
target stimuli and count them silently. Our findings confirmed that the brain processes the target
and non-target stimuli in two different ways. The first, more general processing, could represent
a basic analysis of the external stimuli regardless of their specific meaning and includes e.g.
sensory processing, attention, memory and decision making. The evidence comes from the
findings of identical early and late ERP components evoked after the target and non-target
stimuli in various brain structures, e.g. cingulate gyrus, parahippocampal gyrus, superior,
middle and inferior temporal gyri, hippocampus, and frontoorbital cortex. The second type of
processing takes into account the association between the stimulus and the required response
or the salience of the stimulus. Electrophysiological correlates of such processes can be
represented by intracerebrally recorded cognitive ERP components, e.g. the P3-like waveform,
that can be associated with cognitive, executive and control functions. Our analysis of the
temporal relationship of the P3-like waveform to the stimulus and motor response demonstrated
that this component is a heterogeneous phenomenon. It might reflect stimulus evaluation,
executive functions or attentional and integrative functions. Our next study analysed
7
the temporal characteristics of large negative ERP generated in the human hippocampus during
a simple sensorimotor task. We demonstrated that the hippocampal cognitive potential
with latency of about 420 ms occurs independently of motor execution and within the context
of situation it is most likely related to the evaluation of stimulus meaning.
Our next study aimed to localize neuronal populations that exhibit changes of non-phase
locked EEG oscillatory activity after a presentation of stimuli during a cognitive task.
The significant changes in alpha and beta frequency bands were identified in different frontal
and temporal areas without a specific relationship to a particular structure. The target stimuli
induced power increase in EEG oscillations in alpha frequency band. This could be related
to the activation of neuronal networks functionally linked to the motor response. The non-target
stimuli induced power decrease in EEG oscillations in alpha and in beta frequency bands that
could reflect inhibition processes associated with refraining from motor response. It is evident
that identified power changes are locked to the subject`s response but their functional meaning
is not known yet.
We also utilised the intracerebral EEG data to map the distribution of the ERPs related
to the activation of the brain’s performance monitoring system. We found generators of the
potentials elicited after erroneous responses during Go/NoGo task in various cortical areas
of frontal and temporal lobes, e.g. anterior cingulate gyrus (ACC), medial frontal gyrus,
dorsolateral prefrontal cortex (DLPFC), orbitofrontal and lateral temporal cortices. Our
findings demonstrated a different involvement of the rostral and caudal ACC in error
processing. A follow-up study aimed to localise late postperformance ERPs evoked after correct
responses to the oddball task. We found two functionally distinct neural networks within the
examinand brain regions. While the first network responded only to the target stimuli, the
second network responded independently of a stimulus type. Furthermore, we were able to
identify several sites in ACC, DLPFC, primary motor and lateral temporal cortices that were
simultaneously engaged in evaluation of erroneous responses. Overall, our findings
demonstrated that one anatomical structure can be part of functionally distinct neuronal
networks.
The intracerebral EEG data recorded during the visual oddball task allowed us the
investigation of functional organization of the brain and particularly the investigation of
functional connectivity between the activated brain areas. Correlation analysis of EEG signals
in beta2 frequency band revealed a prominent role of cingulate gyrus in communication
between frontal and temporal brain areas. Using the calculation of “directed transformation
8
function”, we also showed that the directionality of the information flux in frontal lobe during
processing of stimuli associated with active response was from anterior cingulate gyrus to
dorsolateral prefrontal cortex. The findings of these two studies demonstrate the key role of
cingulate gyrus in human behaviour control.
9
Obsah
Seznam použitých zkratek .............………………………………………………………….……….. 10
I. Úvod …..............………………………………………………………………………..………. 11
II. Intracerebrální (intrakraniální) elektroencefalografie ….............…......….…………….……….. 12
1. Typy snímacích elektrod ................................................................................................................13
2. Referenční elektroda ......................................................................................................................17
3. Časové a prostorové rozlišení iEEG ..............................................................................................17
4. Limitace využití iEEG pro výzkumné účely ...................................................................................20
5. Elektrofyziologické koreláty mozkových operací ..........................................................................21
6. Výzkumné užití iEEG ....................................................................................................................24
III. Elektrofyziologické projevy mozkových funkcí studované v průběhu kognitivních úkolů............27
1. Intracerebrální vlna P3-like je heterogenní fenomén ……………………..……..........................29
2. Hipokampální kognitivní potenciál s latencí kolem 420 ms vyvolaný v průběhu
jednoduchého senzorimotorického úkolu nesouvisí s motorickou odpovědí ................................ 32
3. Intracerebrální pozdní ERP komponenty neterčových odpovědí se mohou
ve svém průběhu shodovat s odpovědí terčovou …..................…………………........................35
4. Terčové podněty častěji indukují pokles výkonu EEG aktivity v alfa pásmu, neterčové
podněty naopak častěji nárůst výkonu EEG aktivity v alfa i beta pásmu......................................36
5. Mapování „error“ potenciálů úkolu ukazuje účast řady kortikálních oblastí
v mozkovém systému detekce chybných odpovědí …......................................................….....… 38
6. Po provedení správné odpovědi v rámci oddball úkolu se aktivují dvě funkčně odlišné
neuronální sítě, některé ze zúčastněných struktur jsou současně využívány i při hodnocení
odpovědí chybných ........................................................................................................................40
7. Přední cingulární kortex hraje významnou roli v komunikaci s temporálními
oblastmi mozku …………...............…………………………………………......……................43
8. Při zpracování podnětů spojených s aktivní odpovědí se v rámci frontálního laloku
informace šíří směrem z gyrus cinguli do dorzolaterálního prefrontálního kortexu ......................44
IV. Závěr ..…................…………….……………………………………………………………….. 47
V. Přílohy..……..................…...............................……….......………….…………….………........ 48
VI. Seznam citované literatury ......................................……….…………..……………………......124
VII. Seznam recenzovaných publikací autora .....................................................................................135
10
Seznam použitých zkratek
ACC anterior cingulate cortex (přední cingulární kortex)
CNS centrální nervový systém
CRN correct-related negativity
DLPFC dorsolateral prefrontal cortex (dorzolaterální prefrontální kortex)
DTF directed transformation function
EEG elektroencefalografie, elektroencefalogram
ERD event-related desynchronization
ERN error-related negativity
ERPs event-related potentials
ERS event-related synchronization
FFT fast Fourier transform (rychlá Fourierova transformace)
iEEG intracerebrální (intrakraniální) elektroencefalografie
fMRI functional magnetic resonance imaging (funkční magnetická rezonance)
LFPs local field potentials
MEG magnetoencefalografie
MRI magnetic resonance imaging (magnetická rezonance)
P3 vlna P3
P3-like intracerebrální vlna P3
SEEG stereoelektroenecefalografie
SMA suplementární motorická area
11
I. Úvod
Při studiu mozku a chování se používá řada vyšetřovacích technik, např.
elektroencefalografie (EEG), magnetoencefalografie (MEG) nebo funkční magnetická
rezonance (fMRI). Elektroencefalografie má ve zmíněné skupině technik zvláštní postavení,
protože umožňuje zaznamenávat jeden ze základních projevů činnosti nervového systému, tedy
elektrickou aktivitu mozku. Vedle rutinního skalpového způsobu snímání lze elektrickou
aktivitu neuronů a neuronálních populací registrovat také přímo pomocí elektrod
implantovaných do mozku člověka. V takovém případě hovoříme o intracerebrální
elektroencefalografii (iEEG; někdy se používá obecnější název intrakraniální EEG, který
označuje i snímání subdurálními nebo intrakortikálními elektrodami). Metodu iEEG zavedl
do klinické praxe v České republice v 90. letech minulého století prof. MUDr. I. Rektor, CSc.
z 1. neurologické kliniky u Sv. Anny v Brně. Dodnes je na tomto pracovišti využívána
za účelem lokalizace epileptického ložiska u pacientů s farmakorezistentním typem epilepsie
vybraných do epileptochirurgického programu.
Po mém nástupu na Fyziologický ústav LF MU bylo klíčovým momentem volby mé
vědecké orientace navázání spolupráce prof. MUDr. M. Kuklety, CSc. z našeho ústavu s prof.
MUDr. I. Rektorem, CSc. z 1. neurologické kliniky u Sv. Anny v Brně v rámci projektu
„Plasticita řídicích systémů centrálního nervového systému a možnosti jejího ovlivnění
(MSM141100001)“ a následně projektu „Vnitřní organizace a neurobiologické mechanismy
funkčních systémů CNS (MSM0021622404)“, na jejichž řešení jsem se podílel. V rámci obou
projektů probíhaly na neurologické klinice studie schválené etickou komisí Masarykovy
univerzity, ve kterých pacienti se zavedenými hlubokými elektrodami v průběhu
diagnostického monitorování souhlasili s tím, že budou provádět různé jednoduché kognitivní
úkoly při současném iEEG nahrávání.
Stal jsem se členem řešitelského týmu, který spolupracoval na některých studiích s prof.
MUDr. M. Brázdilem, PhD. Ze začátku jsme využili možnost analyzovat iEEG data, která prof.
Brázdil nasnímal u skupiny 20 pacientů v průběhu jednoduchého zrakového oddball úkolu.
V dalších letech jsme se vedle analýzy iEEG dat podíleli i na jejich získávání u dalších pacientů
v průběhu jiných kognitivních úkolů, jako např. jednoduchého senzorimotorického úkolu nebo
tzv. Go/NoGo úkolu. Jejich společným jmenovatelem je prezentace jednoho nebo střídavě dvou
typů podnětů, přičemž úkolem subjektu je domluveným způsobem (např. stiskem tlačítka)
reagovat na výskyt pouze jednoho typu podnětu. Uvedené úkoly tak nabízejí jedinečnou
možnost studovat intracerebrální elektrofyziologické koreláty neuronálních procesů
12
souvisejících např. se senzorickou diskriminací podnětů, hodnocením významu podnětů,
pamětí, rozhodováním, přípravou a provedením motorické odpovědi. Zmiňované funkce jsou
podkladem schopnosti člověka odpovědět dohodnutým způsobem na definovaný podnět.
Všechny uvedené práce předloženého souboru vycházejí z analýz iEEG registrovaného
v průběhu kognitivních úkolů u epileptických pacientů a jsou vždy výsledkem týmové
spolupráce. Osobně jsem se různou měrou podílel na získávání, analýze dat a publikovaní
výsledků, které v habilitační práci uvádím se souhlasem spoluautorů. V komentovaném
souboru osmi prací jsem u čtyř z nich (příloha 1,2,4 a 6) první nebo korespodující autor, u
zbývajících pak jako spoluator.
II. Intracerebrální (intrakraniální) elektroenecefalografie
S intracerebrálním EEG se poprvé setkáváme již ve 30. letech 20. století (Jaspers a
Carmichael, 1935) relativně brzy po prvním použití skalpové elektroencefalografie u lidí
(Berger, 1929). Za účelem diagnostickým a terapeutickým byla iEEG do klinické praxe
zavedena ve 40. letech 20. století. V popředí zájmu byla v té době bazální ganglia (Meyers a
Hayne, 1948; Knott et al., 1950) nebo thalamus (Ishikawa, 1957). Od 50. let se hluboké EEG
začalo využívat u pacientů s epilepsií (Penfield a Baldwin, 1952; Rasmussen a Jasper, 1958),
protože se v té době vylepšila technika umístění elektrod (Talairach et al., 1952, 1958).
K širšímu využití se dostáváme až se vznikem stereotaktického atlasu, který poskytoval
prostorové koordináty pro stereotaktickou exploraci různých struktur (Talairach a Tournoux,
1988; Talairach et al., 1967). Na základě vyšetření několika desítek epileptických pacientů
potom vznikl i atlas intrakraniální elektroencefalografie (Sperling, 1993).
Metodou iEEG je nejvíce vyšetřovaná skupina pacientů s farmakorezistentním typem
epilepsie, méně často pak pacienti s onemocněním bazálních ganglií, těžkými bolestmi anebo
s mozkovým nádorem (Niedermeyer, 2005). U epileptických pacientů jsou elektrody
ponechány v mozku několik dní za účelem lokalizace epileptogenní zóny v rámci přípravy
operační léčby epilepsie. V průběhu hospitalizace pacienti mohou souhlasit, že budou provádět
různé kognitivní úkoly při současném iEEG nahrávání. Na tomto místě je nutné uvést, že
umístění elektrod a trvání diagnostického postupu je určeno výhradně klinickými požadavky
bez jakékoli vazby na výzkumné úkoly. I přesto však získaná data poskytují unikátní pohled
na fungování lidského mozku. Pacienti mohou z účasti ve zmiňovaných úkolech také profitovat,
protože na základě získaných informací je možné určit specifické funkční oblasti mozku
v blízkosti epileptického zdroje, které nemohou být chirurgicky odstraněny právě z důvodů
13
jejich prokázaného funkčního zapojení do klíčových kognitivních procesů. Zkoumání
mozkových funkcí touto cestou navíc umožňuje pochopit, jak mohou kognitivní procesy
souviset s mozkovou aktivitou, podílející se na vzniku záchvatu (Lachaux et al., 2003).
Intracerebrální (intrakraniální) EEG registruje extracelulární elektrickou aktivitu mozku
dvojího typu. Jednak může snímat akční potenciály generované jednotlivými neurony nebo
malým počtem neuronů – „single“ nebo „multi units recording“. Tento typ aktivity odpovídá
výstupním signálům nervových buněk. Druhý typ představují elektrické signály, které jsou
výsledkem činnosti větších populací neuronů v blízkosti snímacího kontaktu. V zahraniční
literatuře se označují termínem „local field potentials“ (LFPs). Předpokládá se, že odrážejí
sumární synaptickou aktivitu na dendritech neuronů, tedy vstupní signály neuronů (Mukamel
et al., 2012; Lachaux et al., 2003).
II.1. Typy snímacích elektrod
Při registraci intrakraniální EEG se používá několik typů elektrod – subdurální,
intrakortikální a hluboké (Marusič et al., 2006). Subdurální elektrody se implantují pod tvrdou
plenu a umísťují se nejčastěji nad konvexitou hemisfér, případně mohou zasahovat i na jejich
bazální plochu. Vyrábějí se ve dvou variantách, jako tzv. gridy nebo stripy (obr.1).
Obr. 1 Subdurální elektrody: vlevo - grid, vpravo - strip. (zdroj: http://www.diximedical.com)
14
V obou případech se jedná o diskové elektrodové kontakty z neparamagnetického kovu
(platinové případně platino-iridiové) nebo z oceli o velikosti 2-5 mm, které jsou uchycené
v biologicky inertním flexibilním materiálu. V provedení strip jsou kontakty seřazeny do jedné
řady v počtu 4 až 8 na jednom proužku. Ve variantě grid jsou kontakty uspořádány do mřížky
v počtu 8×2 až 8×8. Vzdálenost mezi středy sousedních kontaktů v obou variantách bývá 1 cm.
Výhodou subdurálních elektrod (zejména gridů) je registrace LFPs z relativně velké plochy
hemisfér.
Intrakortikální elektrody jsou v porovnání se subdurálními elektrodami invazivnější,
protože pronikají do kortexu do hloubky několika milimetrů. Registrovaným signálem v tomto
případě je jednotková aktivita neuronů. Snímá se z malého prostoru např. ze 100 kontaktů
při uspořádání elektrod do mřížky v počtu 10×10 na ploše cca 3×3 mm (tzv. „Utah array“,
Hochberg et al., 2006). Jiný způsob uspořádání kontaktů na jedné elektrodě tvaru připínáčku
umožňuje snímání z 22-24 kontaktů o průměru 40 mikrometrů vzdálených od sebe 75-200
mikrometrů (Ulbert et al., 2001a). V souvislosti s prvním typem intrakortikální registrace může
být zajímavé, že s pomocí těchto elektrod dlouhodobě implantovaných do primárního
motorického kortexu kvadruplegického pacienta byl záznam aktivity neuronálních populací
vázaných na zamýšlený pohyb počítačově dekódován a následně úspěšně využit k ovládání
periferních zařízení - pohyb robotickou paží, zapínání televize a otevírání mailů. Jedná se
o první studii dokládající možnosti funkčního použití rozhraní mozek-stroj (Hochberg et al.,
2006).
Hluboké elektrody jsou nejinvazivnější variantou při snímání iEEG, protože pronikají
hluboko do mozkového parenchymu. V české odborné literatuře se také označují termínem
intracerebrální. Využívají se zejména pro registraci z hlubokých mozkových struktur, které
nelze snímat výše uvedenými typy elektrod. Nejčastěji jsou vyšetřovány struktury
meziotemporální – amygdala a hipokampální formace, dále inzulární, cingulární a
frontoorbitální kortex, bazální ganglia, případně thalamus. Z důvodů MRI kompatibility se
používají platinové kontakty v počtu 4 až 18 na jednu jednorázovou semiflexibilní elektrodu.
Průměr elektrody je 0,8 mm, velikost kontaktu je 2 mm a vzdálenost mezi kontakty 1,5 mm
(obr. 2). Používají se různé délky elektrod v závislosti na hloubce uložení cílové vyšetřované
struktury - 5, 10, 15 nebo 18 kontaktů. Elektrody mohou být zaváděny ortogonálně (obr. 3),
tedy z laterálního přístupu kolmo na sagitální rovinu nebo diagonálně, tzn. šikmo z frontálního
nebo okcipitálního přístupu (Marusič et al., 2006).
15
Obr. 2 Nahoře: hluboká elektroda používaná v klinické praxi při registraci iEEG. V elipsovém
výběru je detail koncové části elektrody se třemi kontakty. Dole: schematické znázornění
elektrody. (Zdroj: http://www.diximedical.com).
Vlastní implantaci elektrod předchází získání detailních strukturálním MRI snímků,
které se převádějí do trojrozměrného modelu mozku ve vztahu ke kontrastním značkám a
stereotaktickému keramickému MRI kompatibilnímu rámu. Při následném plánování se hledá
co možná nejpřesnější pozice elektrody ve vztahu k anatomické struktuře, která má být
vyšetřena. Současně se vybírá cesta implantace s nejmenším rizikem poškození cévních
struktur. Zavedení elektrody provádí neurochirurg pomocí speciální dvojité mřížky upevněné
na stereotaktický rám (Marusič et al., 2006). Po zavedení elektrod se provádí kontrolní MRI k
upřesnění jejich lokalizace. Vlivem nehomogenity magnetického pole na rozhraní elektrodatkáň
je šířka elektrody v tomto zobrazení větší, než je její šířka skutečná (obr. 3).
snímací kontakty (2 mm) izolační část (1,5 mm)
16
Obr. 3 Příklady MRI snímků mozku s hlubokými elektrodami implantovanými ortogonálně
do temporálního laloku. Vlevo: řez horizontální, vpravo: řez koronální. Tloušťka zobrazených
elektrod je na snímcích z důvodu nehomogenity magnetického pole větší než ve skutečnosti,
kdy je její průměr 0,8 mm.
Popsané hluboké elektrody s kontakty o velikosti 2 mm a nízké impedanci registrují
LFPs. Při určité modifikaci však mohou snímat i jednotkovou (nebo vícejednotkovou) aktivitu.
Lze toho například dosáhnout umístěním mikrokontaktů na špičku makroelektrody (Fried et al.,
1999). Jinou možností je tzv. hybridní elektroda, na které se střídají kontakty s nízkou
impedancí s kontakty s vysokou impedancí (Howard et al., 1996).
Synonymem monitorace elektrické aktivity při použití hlubokých elektrod je termín
stereoelektroencefalografie (SEEG), který začali prosazovat stoupenci Pařížské školy
(Bancaud, 1959; Talairach et al., 1952). Tento pojem zdůrazňuje skutečnost, že metoda je
založena na stereotaktických principech a s její pomocí se elektrická aktivita registruje ve 3D
prostoru a navíc současně se skalpovým snímáním EEG (Niedermayer, 2005).
Dle trvání implantace elektrod do nervové tkáně lze snímání elektrické aktivity dělit
na akutní (na operačním sále), krátkodobé (za hospitalizace v nemocnici) a dlouhodobé
(pro rozhraní mozek-stroj). Akutní snímání se provádí během neurochirurgického výkonu a trvá
řádově v minutách, zatímco u hospitalizovaných pacientů se zavedenými elektrodami se
elektrická aktivita snímá kontinuálně řádově ve dnech.
17
II.2. Referenční elektroda
Stejně jako u skalpového EEG, i pro intrakraniální snímání je klíčová co možná nejvíc
neutrální referenční elektroda, vůči které se signál na každém kontaktu měří. Kolísání
elektrického pole v blízkosti referenční elektrody by mělo být minimálně o řád nižší než
kolísání elektrického pole zachycené hlubokými elektrodami a to ve všech frekvenčních
pásmech (Lachaux et al., 2003). Tuto podmínku splňují např. spojené elektrody ušní, které však
na druhou stranu mohou být kontaminovány svalovými artefakty případně očními pohyby.
Minimalizovat vliv externích zdrojů umožňují referenční elektrody umístěné uvnitř lebky.
V praxi se používá tzv. snímání bipolární, kdy se měří napětí mezi sousedními kontakty a
referenční elektroda je tak pro každou dvojici jiná. Druhou možností je použití průměrné
hodnoty ze všech intrakraniálních kontaktů. K určení zdroje elektrické aktivity je pak výhodné
data prohlížet v obou způsobech zapojení.
Poměr signál-šum je u iEEG vyšší než u skalpového EEG, protože mezi registrační
elektrodou a nervovou tkání se nenachází bariéry (kost, kůže, vlasy), které by signály z mozku
oslabovaly. Při použití vhodné referenční elektrody jsou iEEG data zatížena svalovými
artefakty a pohyby očí minimálně. Díky uvedeným skutečnostem pak tato metoda umožňuje
jako jediná snímat vysokofrekvenční složky elektrické aktivity mozku, které jsou jinými
metodami nedostupné (Lachaux et al., 2012).
II.3. Časové a prostorové rozlišení iEEG
Časové rozlišení iEEG, stejně jako EEG a MEG, je v milisekundách. Je limitováno
pouze vzorkovací frekvencí použitých registračních zařízení. V současné době se dostáváme
k hodnotám 30 kHz, což umožňuje registrovat právě i jednotkovou aktivitu neuronů.
Jednou z dalších výhod intracerebrální registrace v porovnání se skalpovým EEG nebo
jinými technikami je prostorové rozlišení. To je určeno typem snímací elektrody (impedancí a
velikostí kontaktů), jejím umístěním a také vodivostí mozkové tkáně kolem elektrody.
Při použití intrakortikálních elektrod nebo vysokoimpedančních mikroelektrod umístěných
na těle nebo na špičce hlubokých elektrod je možné snímat jednotkovou aktivitu jednotlivých
neuronů nebo vícejednotkovou aktivitu malé skupiny neuronů v bezprostřední blízkosti
kontaktu. Takové extrémní rozlišení se využívá při mapování struktur během
neurochirurgického zákroku. Například u pacientů s Parkinsonovou chorobou se postupně
zavádí elektroda do bazálních ganglií a dle registrované neuronální aktivity se identifikují
oblasti, které mohou být následně cíleně poškozeny nebo naopak chronicky stimulovány, čehož
18
se využívá při tzv. hluboké mozkové stimulaci. Při výzkumu kognitivních funkcí se registrace
jednotkové aktivity na operačním sále využívá zřídka (např. Ojemann et al., 2002; Ulbert et al.,
2001b), protože klade vysoké požadavky jak na technické provedení, tak na trvání a volbu
kognitivního protokolu.
Lokalizovat zdroj iEEG odpovědi někdy nazývaný jako generátor, tedy oblast nervové
tkáně, jejíž koherentní nervová aktivita vytváří měřitelné elektrické pole, je možné díky
specifické prostorové distribuci elektrického potenciálu, který vytváří. Příkladem může být
tzv. zvrat fáze sledovaného potenciálu na sousedních kontaktech (tzv. „phase reversal“,
obr. 4). Zdroj charakteru dipólu bude v tomto případě se spojnicí kontaktů pravděpodobně
orientován rovnoběžně. Jiným příkladem lokalizace zdroje je výrazná změna amplitudy signálu
na sousedních kontaktech (tzv. „steep voltage gradient“, obr. 5).
Obr. 4 Příklad lokalizace generátoru evokované odpovědi zachycené na elektrodě F v průběhu
jednoduchého senzorimotorického úkolu. Vlevo: mezi kontakty F7 a F8 má potenciál opačnou
polaritu, jedná se o tzv. zvrat fáze („phase reversal“, označeno šipkou). Podněty jsou
prezentovány v čase 0 s. Vpravo: MRI řez v rovině horizontální s elektrodou F zavedenou
do přední části frontálního laloku, kontakty F6-F10 v bílé elipse se nacházejí v gyrus frontalis
medius, Brodmannova area 9.
F6
F7
F8
F9
F10
0 0.6 1.2 s
19
Obr. 5 Příklad lokalizace generátoru evokované odpovědi zachycené na elektrodě B v průběhu
jednoduchého senzorimotorického úkolu. Vlevo: mezi sousedními kontakty B1 až B5 lze
pozorovat výrazné změny amplitudy evokované odpovědi, kontakt B3 je pak nejblíže zdroji.
Podněty jsou prezentovány v čase 0 s. Vpravo: MRI řez v rovině koronální s elektrodou B
zavedenou do temporálního laloku, kontakty B1-B4 v bílé elipse se nacházejí v hipokampu,
kontakt B5 v bílé hmotě.
Převážná většina intrakraniálního snímání u lidí využívá subdurálních nebo
nízkoimpedančních kontaktů na hlubokých elektrodách, které snímají místní potenciály pole.
Určit prostorové rozlišení v těchto případech není tak jednoduché, protože každý kontakt
uvedených elektrod snímá sumární elektrickou aktivitu, která pochází z různých zdrojů
v mozku. Síla elektrického pole zdrojů však klesá se čtvercem jejich vzdálenosti od kontaktu.
To nevylučuje teoretickou možnost ovlivnění ze vzdálenějších zdrojů v případě, že tyto generují
velmi silné elektrické pole v porovnání se zdroji bližšími. Lachaux a spol. (2003) však uvádějí,
že vliv jednotlivých zdrojů vzdálenějších více než 1 cm od kontaktu je zanedbatelný.
I když jsou hluboké elektrody zaváděny do mozkové tkáně s milimetrovou přesností,
je prostorová lokalizace zdrojů elektrické aktivity v mozkové tkáni limitována přesností určení
anatomické pozice elektrod po jejich implantaci. V současné době nejčastěji používané MRI
B1
B2
B3
B4
B5
0 0.6 1.2 s
20
zobrazení elektrod in situ umožňuje určit jejich pozici s přesností řádově v milimetrech.
Shrneme-li výše uvedené informace, iEEG snímaná hlubokými elektrodami se srovnává
s prostorovým rozlišením funkční magnetické rezonance a pozitronové emisní tomografie.
Díky vysokému časovému rozlišení však představuje zajímavý nástroj pro funkční mapování
mozku (obr. 6).
Obr. 6 Schematické porovnání časového (osa x) a prostorového (osa y) rozlišení některých
vyšetřovacích technik (upraveno dle Lachaux et al., 2003).
K doplnění seznamu výhod je nutné také zmínit, že díky intracerebrálním elektrodám
zavedeným do lidského mozku je, na rozdíl od téže techniky používané u zvířat, možné
s vysokým časovým a prostorovým rozlišením studovat procesy související s funkcemi
typickými pouze pro člověka, jako je např. řeč, představivost apod.
II.4. Limitace využití iEEG pro výzkumné účely
Na prvním místě je nutno uvést, že vyšetřovanými subjekty nejsou zdraví jedinci, ale
pacienti s různými typy onemocnění. Při analýze intrakraniálních dat získaných nejčastěji
u pacientů s farmakorezistentním typem epilepsie je vždy otázkou, zda jejich mozkové okruhy
nejsou organizovány vlivem patologie nebo medikace jinak, než u zdravé populace. Výsledky
centimetry
neurony
milisekundy minuty
EEG
MEG
SEEG
fMRI PET
21
studií ukazují, že na behaviorální úrovni je většina pacientů schopna kognitivní úkoly provádět
bez větších problémů. Předpokládá se, že vliv patologické tkáně není vázán na okamžiky
prezentace podnětů a může měnit elektrickou aktivitu v průběhu času nezávisle na jednotlivých
úkolech. Proto se při analýze dat vybírají pouze úseky bez zjevné epileptiformní aktivity a
vybírají se signály z těch elektrod, které se nacházejí dále od epileptického ložiska.
Z metodického hlediska je také nevýhodné, že vyšetřovaní pacienti tvoří méně homogenní
skupinu, např. v parametrech věku nebo kognitivních schopnostech provádět požadované
úkoly, než vybraní zdraví dobrovolníci ve studiích skalpových (Mukamel et al., 2012).
Podstatným omezením iEEG je možnost vyšetření pouze limitovaného objemu
mozkové tkáně. Nelze tedy zachytit aktivitu všech částí neuronálních sítí, které se na
experimentálních úkolech podílejí. Tento nedostatek je možné částečně eliminovat tak, že se
porovnávají data iEEG a fMRI získaná např. u stejných pacientů ze stejného úkolu. Je zajímavé,
že např. v oddball úkolu struktury generující vlnu P3 vykazují často zvýšenou aktivaci ve fMRI
(Clark et al., 2000; Stevens et al., 2000). Ukazuje se, že nárůst tzv. „blood-oxygen-level
dependent“ signálu v fMRI se v některých strukturách shoduje se zvýšením výkonu v pásmu
60-200 Hz (Brovelli et al., 2005), v pásmu 40-150 Hz (Lachaux et al., 2007), v pásmu 70-190
Hz (McDonald et al., 2010) nebo v pásmu 30-70 Hz (Privman et al., 2007). Jaký je vztah mezi
metabolickou a elektrickou aktivitou neuronálních populací zatím není zcela jasné.
II.5. Elektrofyziologické koreláty mozkových operací
Jako jedna z metod funkčního mapování lidského mozku umožňuje intrakraniální EEG
studovat neuronální pochody přinejmenším na dvou úrovních. Na jedné úrovni se vědci snaží
identifikovat jednotlivé složky neuronálních sítí, tzn. hledají oblasti mozku, které vykazují
změny elektrické aktivity (ať už jednotkové nebo LFPs) např. v průběhu kognitivních úkolů.
Na druhé úrovni se studují dynamické vztahy v neuronálních sítích, tzn. časově-prostorová
aktivace a synchronizace jednotlivých zúčastněných oblastí. V obou úrovních se studované děje
odehrávají v určitých frekvenčních pásmech, jejichž označení nejčastěji vychází z klasické
EEG klasifikace jako delta, theta, alfa, beta nebo gama pásma. Někdy se lze setkat s přesnějším
dělením, např. na horní, dolní-1 a dolní-2 alfa pásmo (Klimesch, 2000). Protože se u různých
autorů hranice jednotlivých pásem mohou lišit, užívá se vedle výše uvedeného také rozdělení
s označením číselným, např. frekvenční pásmo 15-40 Hz, 8-15 Hz apod.
Při studiu kognitivních funkcí s pomocí intrakraniálních elektrod se používají stejné
postupy analýzy, jako u registrace ze skalpu. Jednou z nich je zprůměrování EEG odpovědí
22
časově vázáných na opakovanou prezentaci stejného podnětu nebo na provedení jednoduché
akce. Výpočet průměrné amplitudy EEG odpovědi umožňuje studovat fázově koherentní složku
signálu („phase-locked“), tedy evokované děje, které jsou podkladem tzv. „event-related
potentials“ (ERPs). Do češtiny lze přeložit jako „mozkové odpovědi na zaměřené události“
(Stejskal et al., 1993). Pro zkratku a srozumitelnost se však v české odborné literatuře používá
označení ERP nebo ERPs. Ty se skládají z několika komponent – vln (obr. 7), které jsou
definovány latencí, polaritou, distribucí na skalpu a behaviorálními koreláty (např. Halgren et
al., 1998).
Obr. 7 Příklad ERPs po terčových (silná čára) a neterčových podnětech (tenká čára)
registrovaných v průběhu zrakového oddball úkolu z gyrus cinguli anterior vpravo. Podněty
jsou prezentovány v čase 0 s. Z časných ERP komponent je vidět vlnu N1 a P2. Kognitivní vlna
P3 se v intracerebrálním EEG označuje jako P3-like.
(Obrázek z přednášky „Electrophysiological correlates of neuronal processing during cognitive
tasks studied by the intracranial EEG in humans“ prezentované na 25. mezinárodní CIANS
konferenci v Bratislavě, Slovensko, 2016).
Po sobě jdoucí vlny lze rozdělit na časné senzorické potenciály, které jsou zejména
ovlivnitelné fyzikálními vlastnostmi vyvolávajícího podnětu (Jewett et al., 1970; Cracco a
Cracco, 1976) a na pozdní kognitivní potenciály, podmíněné kognitivními procesy (Sutton et
al., 1965; Donchin et al., 1978; Hillyard et al., 1978, Halgren et al., 1998). Mezi ERP
časné senzorické
ERP komponenty
pozdní kognitivní
ERP komponenty
terčová odpověď
neterčová
odpověď
vlna P3-like
N1
P2
23
komponenty časné řadíme vlny s latencí obvykle kratší než 250ms, označované nejčastěji jako
P1, N1, P2, výjimečně některými autory uváděná i N2 (Antal et al., 2000). ERP komponenty
s latencí delší než 250 ms jsou pak označovány jako kognitivní a patří sem např. vlny N2, P3,
N400. Použité dělení je však pouze orientační, protože přesná hranice mezi senzorickými a
kognitivními potenciály není doposud jednoznačně určena. Na tomto místě je zapotřebí ještě
uvést, že polarita jednotlivých komponent evokovaných odpovědí u intracerebrálních dat nemá
stejnou výpovědní hodnotu, jako při registracích na skalpu. Navíc není také doposud jasné,
jakým způsobem se v mozkové tkáni identifikované generátory ERP podílejí na skalpových
signálech. Detailnější popis problematiky ERPs lze nalézt v dizertační práci (Roman, 2004).
Interpretace ERPs vychází z předpokladu, že při opakované prezentaci stejného podnětu
se shodně mění i neuronální aktivita ve stejných anatomických oblastech a ve stejném časovém
okně. Někteří autoři popisují tuto skutečnost jako „phase resetting“ (Brandt, 1997).
Zprůměrněním takto vyvolaných EEG odpovědí se pravidelná aktivita zvýrazní, zatímco
neuronální aktivita, která není vázána na prezentaci podnětu (šum), se potlačí. Tento
jednoduchý matematický postup je použitelný i tehdy, když evokovaná odpověď při některé
prezentaci podnětu chybí nebo, když její latence od podnětu není konstantní. Takové kolísání
však musí být mnohem menší než trvání studovaného evokovaného potenciálu. Z uvedeného
pak vyplývá i jisté omezení. Např. tímto postupem nelze detekovat evokované potenciály
o frekvenci vyšší než 30 Hz, pokud jednotlivé odpovědi nejsou s podnětem synchronizovány
s milisekundovou přesností (Lachaux et al., 2003).
Prezentace podnětů však vedle evokované odpovědi může vyvolat takové změny LFPs,
při kterých jednotlivé EEG oscilace nejsou fázově koherentní. Hovoříme o tzv. odpovědi
indukované. V tomto případě se analyzuje především změna amplitudy nebo výkonu oscilací,
nikoliv fáze ("time-locked; nonphase-locked"), která se navíc obvykle objevuje s delším
časovým odstupem od okamžiku stimulace než odpověď evokovaná. Předpokládaným
podkladem tohoto typu EEG odpovědi je změna počtu aktivovaných oscilátorů (tj. oscilačních
neuronálních celků) v daném frekvenčním pásmu (Klimesch, 1999).
U indukovaných odpovědí vyhodnocujeme především relativní změny amplitudy
oscilací EEG signálu ve vymezeném frekvenčním pásmu vzhledem k amplitudě signálu
v krátkém časovém okně před prezentací podnětu (Pfurtscheller, 1992). Nárůst amplitudy je
označován ERS („event related synchronization“) a pokles ERD („event related
desynchronization“). Pro analýzu indukovaných změn se používá výpočet amplitudových nebo
výkonových obálek pomocí Hilbertovy transformace s využitím FFT nebo vlnkové
24
transformace. Pro různá frekvenční pásma lze vytvářet časově-frekvenční mapy. Jednotlivé
mapy nebo obálky se průměrují po jednotlivých EEG epochách („EEG trials“).
Pro analýzu časově-prostorové aktivace zúčastněných neuronálních oblastí, tzn. míry
jejich funkčního vztahu – tzv. funkční konektivity, se např. používá výpočet korelace signálů
ze dvou a více míst mozku v časové oblasti (korelace; např. Kukleta et al., 2009) nebo
ve frekvenční oblasti (koherence výkonových spekter; Rappelsberger et al., 1993). Kromě
lineární korelace a koherence patří v současné době mezi nejpoužívanější metody analýzy
funkční konektivity i metody založené na výpočtu okamžité fáze signálu, nelineární a kauzální
metody. Z metod využívajících okamžitou fázi signálu lze uvést například „Phase
Synchronization“, „Mean Phase Coherence“ (Mormann et al., 2000) nebo „Phase-lag Index“
(Stam et al., 2007). Kromě zachycení okamžitých fázových změn mezi dvojcí EEG signálů mají
tyto postupy schopnost lépe potlačit společné, nežádoucí složky v signálech. Z nelineárních
metod lze použít např. h2 – nelineární korelaci. I přesto, že je použití nelineárních analýz
v signálech EEG z mnoha důvodů diskutabilní, dosahuje tato metoda dobrých výsledků,
především v kontextu dynamických změn výrazných behaviorálních stavů, jako jsou například
epileptické záchvaty (Wang et al., 2014, Jirsa et al., 2014).
Pro posouzení kauzality zkoumaných dějů v EEG, mohou výše zmíněné konvenční
metody nabídnout pouze omezenou informaci. Pro popis časových posloupností v EEG
signálech je proto často používána Grangerova kauzalita, která na základě predikce a výsledné
chyby určuje, který EEG signál kauzálně ovlivňuje průběh signálů okolních (Granger, 1969).
V případě zkoumání kauzální konektivity (tzv. efektivní konektivity) mezi více než dvěma EEG
signály jsou využívány multivarietní metody založeny na principu Grangerovy kauzality –
„Directed Transfer Function“ a „Partial Directed Coherence“ (Baccala a Sameshima, 2001).
Při analýzách konektivity však obecně platí, že hodnota výsledných parametrů je významně
závislá na strukturálních spojích mezi odpovídajícími anatomickými oblastmi a je ovlivněná
přítomnou patologií (Tucker et al., 1986).
II.6. Výzkumné využití iEEG
Výzkumné využití iEEG u lidí představuje minoritní podíl mezi moderními technikami
funkčního mapování mozku z důvodů vysoké náročnosti na její provedení a relativně malého
počtu pacientů, kteří z klinických důvodů tuto exploraci podstoupí. K náročnosti do jisté míry
přispívá nutnost multidisciplinárního přístupu - většinou se na výzkumu podílejí odborníci
25
z oboru neurologie, neurochirurgie, neurofyziologie, neuropsychologie a kognitivních
neurověd. I přes uvedené skutečnosti však tato metoda poskytuje unikátní data díky její
milimetrové a milisekundové rozlišovací schopnosti.
V následujících odstavcích jsou uvedeny některé zajímavé nálezy humánních iEEG
studií, které přispěly k rozšíření našich znalostí v různých oblastech výzkumu mozku (řeč,
motorika, percepce, emoce a paměť). Širší výčet prací lze nalézt v přehledových článcích
Lachaux a spol. (2003), Jacobs a Kahana (2010) nebo Mukamel a Fried (2012).
Při výzkumu řeči a řečových funkcí Penfield a Roberts (1959) a později Ojemann a spol.
(1989) prokázali účast mnoha kortikálních oblastí v řečově dominantní hemisféře pomocí
elektrické stimulace kortexu během neurochirurgických zákroků. Sahin a spol. (2009)
na základě registrace potenciálů místního pole zase ukazují, že lexikální, gramatické a
fonologické procesy řeči jsou v Brokově oblasti zpracovávány v odlišných časových obdobích.
Dalším zajímavým nálezem je, že např. premotorické oblasti během mluvení modulují
neuronální aktivitu ve sluchovém kortexu (Greenlee et al., 2011).
Somatotopická organizace primárního motorického i senzorického kortexu byla
prokázána pomocí elektrické stimulace kortexu (Penfield a Boldrey, 1937) a také sledováním
výkonu v alfa, beta a gama pásmech pro pohyby různých částí těla (Crone et al., 1998a,b).
S využitím stimulace nervus medianus se Allison a spol. (1989) podařilo zmapovat časově
prostorovou reprezentaci odpovědi na tuto stimulaci v odpovídajícím senzorimotorickém
kortexu. Fried a spol. (1991) jako první prokazují somatotopickou organizaci lidské
suplementární motorické oblasti (SMA). Navíc elektrickou stimulací této oblasti spustili
u pacientů potřebu provést pohyb, demonstrujíce tak spojení mezi SMA a volním pohybem.
Podobný vztah k volnímu řízení pohybu nalézá Desmurget a spol. (2009) i u dolního
parietálního kortexu, kdy jeho stimulace vedla k vyvolání pocitu, že pacient by měl provést
pohyb. Při vyšší intenzitě stimulace dokonce došlo k navození iluze, že pohyb byl proveden.
Nezávisle na vůli pacienta však skutečný pohyb spouští stimulace v premotorické korové
oblasti (area 6). Tankus a spol. (2009) demonstrují, že neuronální aktivita SMA oblasti je
nepřímo úměrná rychlosti prováděného pohybu. Odpovědi neuronů téže oblasti byly
registrovány i při pouhé představě pohybů (Amador a Fried, 2004).
V lidském hipokampu a parahipokampálním gyru Ekstrom a spol. (2003) jako první
demonstrují existenci „place“ a „view“ neuronů, jejichž aktivita je podkladem prostorové
orientace. V jiné studii Allison a spol. (1999) identifikovali komponenty ERPs související
s percepcí lidských tváří a podrobně zmapovali jejich lokalizaci v okcipitotemporálním
26
kortexu. V jiné studii byly meziotemporálně identifikovány takové neurony, které se aktivovaly
jak při prohlížení zrakových podnětů, tak při jejich představování při zavřených očích (Kreiman
et al., 2000a,b). Jiná studie ukázala, že stupeň fázové shody jednotkové aktivity s pomalými
oscilacemi místních potenciálů pole v pásmu 3-8 Hz v hipokampu a amygdale je ukazatelem
výkonnosti paměti (Rutishauser et al., 2010).
Dále při výzkumu emocí se např. ukázalo, že aktivita amygdaly se při rozlišování
averzivních, neutrálních a pozitivních podnětů liší jak v indukované (Oya et al., 2002; Sato et
al., 2010) tak v evokované odpovědi (Krolak-Salmon et al., 2004). Pocity znechucení byly
vyvolány elektrickou stimulací těch oblastí inzuly, které vykazují nárůst amplituy ERP při
pozorování výrazů znechucení (Krolak-Salmon et al., 2003). Další studie mapují lokalizaci tzv.
zrcadlových neuronů, které se aktivují v případě, že vyšetřovaný subjekt sám provádí nějakou
činnost nebo stejnou činnost pozoruje u ostatních. Jejich existence byla prokázána nejen
v motorickém a senzorickém kortexu, ale i v SMA nebo hipokampu (Mukamel et al., 2010a).
Analogicky aktivní neurony lze nalézt i v předním cingulárním kortexu u subjektů v situacích,
kdy sami jsou vystaveni bolestivé stimulaci nebo když pozorují jiné osoby vystavené stejné
bolestivé stimulaci (Hutchison et al., 1999).
27
III. Elektrofyziologické projevy mozkových funkcí studované v průběhu kognitivních
úkolů.
V šesti předkládaných studiích (příloha 1,3,4,6,7,8) jsme zpracovávali prof. Brázdilem
dříve nahraná intracerebrální EEG data, která byla získána v průběhu jednoduchého zrakového
oddball úkolu u souboru 20 epileptických pacientů (ve věku od 23-45 let). V dalších studiích
byl použit jednoduchý senzorimotorický úkol (příloha 2; vyšetřeno 11 epileptických pacientů,
průměrný věk 35±11 let) a tzv. Go/NoGo úkol (příloha 5; vyšetřeno 7 epileptických pacientů,
průměrný věk 29±4 roky), jejich detailnější popis je uveden v komentáři k dané publikaci.
Ve všech studiích byly při registraci EEG signálu použity standardní semiflexibilní
vícekontaktové elektrody ALCIS o průměru 0,8 mm. EEG bylo nahráváno 128 kanálovým
TrueScan EEG systémem na vzorkovací frekvenci 1024 Hz.
Studované elektrofyziologické koreláty jsou generovány v různých frekvenčních
pásmech. Z toho důvodu jsou EEG data v každé studii po odstranění artefaktů a před vlastní
segmentací filtrována offline ve frekvenčních pásmech odpovídajících studovanému korelátu.
Následně pak do analýzy vstupují EEG přeběhy (synonyma: EEG segmenty, EEG epochy,
„EEG trials“) v trvání několika stovek milisekund před a po okamžiku prezentace podnětů nebo
registrované odpovědi.
EEG data získaná při současné registraci ze skalpové CPz elektrody, umístěné
v poloviční vzdálenosti mezi elektrodami Cz a Pz mezinárodního 10-20 systému, byla v našich
studiích málo využitelná z důvodů nižšího poměru signál/šum a použitých analýz. Ojedinělé
nálezy z této elektrody zmiňujeme v následujících komentářích jen výjimečně.
Pro analýzu iEEG dat jsme ve většině případů použili software ScopeWin, jehož
autorem je ing. Pavel Jurák, CSc, případně software Physioplore, jehož autorem je ing. Jan
Chládek, PhD., oba autoři jsou zaměstnanci Ústavu přístrojové techniky Akademie věd
ČR, Brno.
Oddball úkol
V základní variantě oddball úkolu (jednoduchý rozhodovací úkol; pro anglické slovo
„oddball“ nemáme v češtině překlad) jsou náhodně prezentovány dva typy senzorických
podnětů stejné modality – terčové (synonymum vzácné; angl. „rare, target, deviant“) a
neterčové (synonymum časté; anglicky standard, frequent, non-target“) v poměru 1:5 (obr. 4).
Mezi jednotlivými podněty se nachází krátká pauza (cca 2-5 s). V našem případě zrakové
stimulace s použitím písmen jako podnětů představovalo písmeno „X“ podněty terčové,
28
písmeno „O“ podněty neterčové. Pokusná osoba byla instruována reagovat na výskyt každého
terčového podnětu stlačením ručního spínače a v duchu tyto podněty počítat, zatímco ostatní
podněty nechávat bez povšimnutí (např. Polich, 2007).
Obr. 4 Schematické znázornění jednoduchého oddball úkolu. V naší studii bylo terčovým
podnětem písmeno „X“ a neterčovým podnětem písmeno „O“ (horní záznam). Na výskyt
terčového podnětu pokusná osoba reagovala stiskem tlačítka (dolní záznam). Oba záznamy jsou
vždy nahrávány do přídavných kanálů zesilovače současně s kanály EEG.
Oddball úkol je do jisté míry značně jednoduchý (percepce dvou podnětů stejné
modality) a tudíž velmi snadno použitelným pro základní i aplikovaný neurofyziologický
výzkum s využitím různých vyšetřovacích metod. Na druhé straně však určení významu
podnětů a jejich randomizovaný výskyt v průběhu úkolu vyžaduje zapojení mentálních operací,
které jsou hlavním cílem zkoumání v této oblasti neurofyziologie a při kterých se angažuje řada
kortikálních a subkortikálních struktur. S využitím moderních vyšetřovacích metod s lepším
prostorovým i časovým rozlišením, jednou z nich je právě intracerebrální elektroencefalogarfie,
je možné získat detailnější informace o zúčastněných neuronálních sítích a časo-prostorových
vztazích mezi neuroanatomickými strukturami, které jsou aktivovány v průběhu základních
mentálních operací.
29
III.1. Intracerebrální vlna P3-like je heterogenní fenomén
V naší první studii (příloha 1) jsme se zaměřili na detailnější analýzu kognitivní
komponenty ERP označované jako vlna P3, konkrétně na její intracerebrální variantu
označovanou jako vlna P3-like. V oddball úkolu je typicky přítomná po terčových podnětech.
Její další název je vlna P300, což souvisí s průměrnou vrcholovou latencí této komponenty
(kolem 300 ms) u zdravých jedinců při použití sluchové modality stimulace. Latence vizuální
vlny P3 se pohybuje v rozmezí 250-600 ms od začátku podnětu a může se měnit v závislosti
na typech podnětů, složitosti úkolu a na vyšetřovaných osobách (Comerchero a Polich, 1998).
Časově se však často překrývá s okamžikem pohybu, může jej předcházet nebo následovat, a
proto na pohyb-vázané potenciály mohou zkreslit její skalpovou distribuci a změnit její latenci
(Jentzsch a Sommer, 2001; Salisbury et al., 2001; Brázdil et al., 2003).
I přes nejednoznačné nálezy ve studiích zaměřených na analýzu vztahu mezi latencí
vlny P3 a reakční dobou (např. Gehring et al., 1992; Mecklinger et al., 1992; Gunter et al.,
1994), považuje většina autorů vlnu P3 obecněji za korelát procesů souvisejících
s vyhodnocováním podnětů („stimulus-related“) a nesouvisejících s procesy výběru a
provedením odpovědi (Pfefferbaum et al., 1986). Tento kognitivní potenciál bývá tedy
interpretován jako měřítko rychlosti zařazení nebo vyhodnocení podnětu (Polich, 1986), jako
ukončení poznávacího procesu v mozku („cognitive closure“), či jako korelát aktualizace
paměťových funkcí před další akcí nebo bývá spojován s mechanismy pozornosti a orientace
(Paller et al., 1987).
Ve studiích využívajících intracerebrálních elektrod u epileptických pacientů byly
nalezeny generátory vlny P3 v různých strukturách (Baudena et al., 1995; Halgren et al.,
1995a,b; Brázdil et al., 1999). Byly popsány dva charakteristické, na modalitě podnětu
nezávislé, intracerebrální vzorce vlny P3. První představuje trifázická vlna s ostrými vrcholy
(negativní–pozitivní–negativní) označovaná jako N2a-P3a-SW. Bývá spojována se systémem
zaměřené pozornosti a byla pozorována opakovaně v těchto oblastech: gyrus temporalis
medius, gyrus fusiformis, gyrus parahipokampalis, gyrus cinguli, dorzolaterální prefrontální
kortex a dolní parietální kortex. Druhým vzorcem je široká vlna P3b, která je součástí vlny
N2b-P3b a která je spojována se systémem „kódujícím události“, byla nalezena např.
v laterálním orbitofrontálním kortexu, amygdale a hipokampu (Halgren et al., 1995a,b;
Brázdil et al., 1999).
30
Na rozdíl od všech do té doby známých intrakraniálních studií zabývajících se lokalizací
kognitivních potenciálů v mozku jsme se v této studii zaměřili na časový vztah intracerebrální
vlny P3-like k podnětu a motorické odpovědi. Využili jsme přitom skutečnosti, že 1) EEG
záznamy z hlubokých elektrod mají vyšší poměr signál/šum a tudíž stačí zprůměrnit menší
počet EEG přeběhů k vykreslení studovaného potenciálu; 2) u každého EEG přeběhu jsou vedle
přesné časové identifikace prezentace podnětu ještě zaznamenány okamžiky stisku tlačítka; 3)
reakční doba, tj. čas od prezentace podnětu po okamžik stisku tlačítka, není v jednotlivých EEG
přebězích u každého subjektu stejná a individuálně je tak možné rozdělit EEG přeběhy
do podskupin s krátkou a dlouhou reakční dobou.
Příloha 1. Roman R, Brazdil M, Jurak P, Rektor I, Kukleta M. Intracerebral P3-like waveforms
and the length of the stimulus-response interval in a visual oddball paradigm. Clinical
Neurophysiology 2005;116:160-171.
K analýze jsme použili intracerebrální data registrovaná v průběhu zrakového oddball
úkolu u 17 epileptických pacientů (4 ženy a 13 mužů). K odstranění vlivu vysokofrekvenčních
složek EEG signálu, které nepřispívají ke generování vlny P3-like, jsme EEG data nejdříve
digitálně filtrovali (pásmová propust 0,5-5,5 Hz). U každého pacienta bylo zprůměrněno
přibližně 50 přeběhů po terčových podnětech a 150 přeběhů po podnětech neterčových. Vlna
P3-like v rozsahu latencí 250-600 ms byla identifikována ve 180 místech frontálního,
temporálního a parietálního laloku. V dalším kroku jsme EEG přeběhy po terčových podnětech
rozdělili do dvou podskupin s kratší a delší reakční dobou. Přeběhy (počet 16±3)
v podskupinách byly poté zprůměrněny dvěma způsoby a to od podnětu a od stisku spínače.
Porovnání latencí vln P3-like v obou podskupinách při dvou způsobech zprůměrnění
nám umožnilo rozdělit vlny P3-like na tři typy s difúzní intracerebrální distribucí. Jednotlivé
typy nebyly specifické pro určité struktury a u většiny pacientů jsme mohli identifikovat více
než jeden případ P3-like vlny každého typu v různých strukturách.
Prvním typem byla vlna časově vázaná na podnět (průměrná latence 393 ms).
Zprůměrněním od podnětu vykazovala tato vlna stejnou latenci v obou podskupinách.
Zprůměrněním od stisku spínače se však latence lišily (delší byla v podskupině s delší reakční
dobou) rozdílem, který představoval průměrně 91,1±22,9 % odpovídajícího rozdílu v průměrné
reakční době obou podskupin. Takový typ potenciálu může představovat procesy související se
senzorickou diskriminací a vyhodnocením podnětu. Může být podkladem skalpových P3 vln
generovaných po „distraktorech“ (Knight, 1998), registrovaných v pasivním oddball úkolu
31
(Polich, 1989) nebo P3 vln, jejichž latence se měnila s reakční dobou při snižování intenzity
podnětů (Verleger et al., 1991) nebo při snižování diskriminability podnětu (Novak et al., 1990).
Druhým typem byla vlna časově vázaná na odpověď (průměrná latence 440 ms).
Stejnou latenci v obou podskupinách vykazovala tato vlna při zprůměrnění od stisku spínače.
Při zprůměrnění od podnětu se však latence lišily (byla delší v podskupině s delší reakční
dobou) rozdílem, který představoval průměrně 95,1±20,1 % odpovídajícího rozdílu v průměrné
reakční době obou podskupin. Taková vlna P3-like pravděpodobně reprezentuje procesy
související s vytvořením vnitřní reprezentace odpovědi, její přípravou, zahájením a vlastním
provedením. Její existence by mohla např. vysvětlit, proč v úkolech s minimálními požadavky
na vyhodnocení podnětu vedlo provedení odpovědi k ovlivnění latenc skalpové vlny P3
(Doucet a Stelmack, 1999) nebo proč změnilo její skalpovou distribuci (Salisbury et al., 2001).
Nejčastěji jsme pozorovali třetí typ vlny bez jasné časové vazby na podnět a odpověď
(průměrná latence 425 ms). U tohoto typu vlny představoval rozdíl latencí mezi oběma
podskupinami při zprůměrnění od podnětu 57,4±26,0 % a při zprůměrnění od odpovědi
51,9±18,8 % rozdílu v průměrné reakční době obou podskupin. Zdá se, že poslední typ vlny
P3-like by mohl souviset s paměťovými a pozornostními operacemi nebo procesy integrujícími
zpracování podnětu a provedení odpovědi, tzn. rozhodnutí o odpovědi na základě analýzy
senzorické informace. S těmito operacemi bývá asociována skalpová vlna P3b
s centroparietálním maximem (McCarthy et al., 1997). Redukce vlny P3b u pacientů
s temporoparietální lézí je doprovázena deficity pozornosti a paměti (Woods et al., 1993).
Nálezy třetího typu P3-like vlny v hipokampu, parahipokampálním gyru, amygdale a na dalších
místech mohou podpořit výše uvedené zjištění. Je pravděpodobné, že rozhodování, pozornostní
a paměťové procesy nemusí být přesně časově vázány na okamžik prezentace podnětu a jejich
časová vazba může kolísat v závislosti na okamžitých podmínkách. Stejné vysvětlení může být
použito pro procesy počítání terčových podnětů, které pak mohou přispívat ke vzniku třetího
typu vlny P3-like.
Výsledky naší studie tedy prokázaly, že vlna P3-like vyvolaná v průběhu zrakového
oddball úkolu je heterogenní fenomén odrážející více mozkových funkcí. Zajímavým vedlejším
nálezem bylo zjištění, že oba intracerebrální vzorce – trifázická vlna N2a-P3a-SW i široká vlna
P3b – byly nalezeny u všech tří typů P3-like vln. Z uvedeného vyplývá, že tyto vzorce
evokované EEG odpovědi registrované přímo v mozkové tkáni odrážejí spíše určitou
elektrofyziologickou zákonitost než specifické kognitivní operace.
32
III.2. Hipokampální kognitivní potenciál s latencí kolem 420 ms vyvolaný v průběhu
jednoduchého senzorimotorického úkolu nesouvisí s motorickou odpovědí
V práci Roman a spol. (2005) zaměřené na analýzu vlny P3-like vyvolané v oddball
úkolu jsme pozorovali, že se v hipokampu člověka pravidelně vyskytuje významný pomalý
ERP s latencí kolem 450 ms. Stejný nález při použití téhož úkolu popisuje také jiná skupina
(Ludowig et al., 2010). Potenciál velmi podobného charakteru byl v hipokampu nalezen také
při použití jiných experimentálních úkolů, např. při studiu paměťových procesů (Paller a
McCarthy, 2002). Pro svoji přítomnost v řadě humánních studií (např. Grunwald et al., 1995;
Brázdil et al., 2001; Fell et al., 2005; Boutros et al., 2008; Axmacher et al., 2010) je tento
potenciál chápán jako korelát kognitivních procesů zprostředkovaných hipokampem v různých
kognitivních úkolech. V některých studiích byl tento potenciál považován za jeden z generátorů
skalpové vlny P300 a proto jej někteří autoři označují např. jako MTL-P300 (Fell et al., 2004),
jako hluboká P3b (Halgren et al., 1995b) nebo P3-like vlna (Brázdil et al., 2003). Protože tento
hipokampální ERP není vázán na některé funkce asociované s funkcemi vlny P300, Paller a
McCarthy (2002) navrhují používat popisný termín „large negative ERP“.
Výsledky většiny intracerebrálních studií prokazují účast hipokampu v kognitivních
procesech, jako jsou např. hodnocení významu podnětů (Rosburg et al., 2007; Grunwald et al.,
1998; Strange a Dolan, 2001) a při paměťových úkolech - prostorová paměť (např. Watrous et
al., 2011), kódování a vybavování epizodické paměti (např. Fernández el al., 2002; Amaral a
Lavenex, 2007; Eichenbaum et al., 2012). Některé animální studie však ukazují, že
hipokampální theta aktivita je spojena s motorickým chováním (Vanderwolf, 1969; Wyble et
al., 2004; Shin, 2011). Intracerebrální studie u člověka také v hipokampu prokazují na pohyb
vázaný nárůst theta oscilační aktivity (Ekstrom et al., 2005). Tyto nálezy jsou v souladu se
senzorimotorickou integrační teorií, která předpokládá, že hipokampální theta aktivita odráží
jak zpracování senzorických informací souvisejících s přípravou pohybu tak jeho provedením
(Bland a Oddie, 2001).
Další doklady pro účast hipokampu při provádění pohybu pocházejí ze studií, které
ukazují změny hipokampální evokované aktivity v průběhu motorické odpovědi. Například
v oddball úkolech, je amplituda hipokampálního pomalého ERP vždy vyšší po terčových
podnětech, vyžadujících stisk tlačítka, v porovnání s odpovědí po neterčových podnětech, kdy
motorická odpověď není požadována (Brázdil et al., 1999; Fell et al., 2005; Ludowig et al.,
2010). Amplituda hipokampálního pomalého ERP po terčových podnětech je také vyšší
v případě, kdy pacient označí přítomnost terčového podnětu stiskem tlačítka v porovnání
33
s mentálním počítáním (Brázdil et al., 2003). Navíc v naší předchozí studii (Roman et al., 2005)
jsme pozorovali, že hipokampální P3-like vlna může být časově vázána jak na okamžik
prezentace podnětu, tak na okamžik motorické odpovědi. Tyto nálezy naznačují, že
hipokampální pomalá evokovaná aktivita může souviset vedle vyhodnocení podnětů a
paměťových procesů také s provedením pohybu.
Abychom objasnili tuto možnou souvislost, zaměřili jsme se v další naší intracerebrální
studii (příloha 2) na analýzu časového vztahu mezi latencí hipokampálního pomalého ERP a
okamžikem motorické odpovědi. Zvolili jsme jednoduchý úkol, při kterém byl prezentován
pouze jeden typ podnětu (1 kHz tón; celkem 150 podnětů; interstimulus interval 4-6s). Pacienti
byli instruováni stisknout tlačítko spínače po zaznění tónu, přičemž nebyl kladen důraz
na rychlost odpovědi.
Příloha 2. Roman R, Brázdil M, Chládek J, Rektro I, Jurák P, Světlák M, Damborská A, Shaw
DJ, Kukleta M. Hippocampal negative event-related potential recorded in humans during a
simple sensorimotor task occurs independently of motor execution. Hippocampus
2013;23(12):1337-1344.
Tato práce získala v roce 2014 ocenění České společnosti pro klinickou neurofyziologii ČLS
JEP - 1.místo v soutěži o nejlepší vědeckou práci za rok 2013.
Z metodického hlediska se jedná o první iEEG studii, ve které byly pro stanovení latence
intracerebrálního ERP použity parametry odvozené z plochy studovaného potenciálu.
Vycházeli jsme z obdobné metodiky doporučené pro analýzu skalpových ERP (Luck, 2005).
V našem případě jsme měřili latenci těžiště plochy potenciálu (tzv. „50% area latency“)
v časovém okně, jehož počáteční a koncový bod ležely na průsečíku potenciálu a pomyslné
čáry, která jej protínala v polovině jeho amplitudy („50% peak amplitude“).
Výhodou tohoto parametru je, že získané hodnoty jsou mnohem méně ovlivněny
šumem. V porovnání s vrcholovou latencí tak přesněji odráží časovou pozici potenciálu
od okamžiku prezentace podnětu a je proto pro analýzu vztahu latence studovaného potenciálu
a reakční doby, konkrétně mediánu reakční doby, vhodnější. Počáteční bod vymezeného
časového okna jsme potom použili i pro určení latence začátku studovaného potenciálu,
tzv. „onset latency“ (obr. 5). Uvedená měření jsme prováděli pomocí software Physioplore,
jehož autorem je ing. J. Chládek, Ph.D. z Ústavu přístrojové techniky Akademie věd ČR, Brno.
34
Obr. 5 Schematické znázornění parametrů odvozených z plochy studovaného potenciálu
(popis parametrů je uveden v textu).
Vyšetřili jsme celkem 11 pacientů (9 žen a 2 muži) s farmakorezistentním typem
epilepsie. Z celkového počtu 91 hlubokých elektrod implantovaných do frontálních,
parietálních a temporálních laloků jsme analyzovali data z 22 elektrod (9 v pravé, 13 v levé
hemisféře) umístěných do hipokampu.
Studovaný pomalý ERP je evokovanou odpovědí v pásmu delta (kolem 3 Hz) a proto
jsme v prvním kroku analýzy EEG data filtrovali (pásmová propust 0,1-5,5 Hz). Vybrané EEG
přeběhy bez artefaktů (průměrný počet 92±21) jsme zprůměrnili od okamžiku prezentace tónu.
U všech vyšetřených hipokampů jsme našli lokálně generovaný pomalý ERP s průměrnou
latencí kolem 420 ms a amplitudou -129±84 µV. Latence těžiště hipokampálního ERP byla
ve 14 případech kratší než medián reakční doby, ve zbývajících 8 případech byla delší než
medián reakční doby.
V dalším kroku analýzy jsme použili originální postup, při kterém jsme EEG přeběhy
rozdělili dle reakční doby do pěti podskupin a odděleně zprůměrnili. U takto získaných
potenciálů jsme opět změřili jejich latenci těžiště, které jsme navíc přepočítali na procentuální
podíl reakční doby a získali jsme tak relativní latenci. Parametr latence a relativní latence jsme
korelovali s reakční dobou měřenou v jednotlivých podskupinách. Výsledky korelační analýzy
prokázaly, že latence hipokampálního pomalého ERP měřená v pěti podskupinách EEG
35
přeběhů nekorelovala s mediánem reakční doby těchto podskupin. Hodnota korelace relativní
latence s mediánem reakční doby naopak dosáhla v 16 z 22 případů signifikantní hladiny
významnosti. Výsledky naší analýzy ukázaly, že hipokampální pomalý ERP u člověka
vyvolaný v průběhu jednoduchého senzorimotorického úkolu je časově nezávislý na okamžiku
provedení motorické odpovědi. Jedná se tedy o elektrofyziologický korelát velmi
pravděpodobně související s vyhodnocováním významu podnětu v kontextu situace.
III.3. Intracerebrální pozdní ERP komponenty neterčových odpovědí se mohou ve svém
průběhu shodovat s odpovědí terčovou
Průběh evokované odpovědi přibližně od 200-250 ms po prezentaci terčových a
neterčových podnětů v oddball úkolech je ve skalpových studiích odlišný. Po neterčových
podnětech je evokovaná změna minimální, pokud vůbec, zatímco po terčových se např.
objevuje významná vlna P3. Tato skutečnost se stala i jedním z kritérií pro identifikaci vlny P3.
Podobný nález lze najít i u intrakraniálních registrací (např. Halgren et al., 1998; McCarthy et
al., 1989; Ludowig et al., 2010). Nepřítomnost evokované odpovědi po neterčovém podnětu je
dávána do souvislosti s tím, že po jeho detekci není vyžadována žádná akce, tj. pokusné osoby
si jej nemají všímat. Z toho důvodu byla evokovaná odpověď po neterčových podnětech při
analýzách ERP opomíjena.
Z podstaty oddball úkolu však lze předpokládat, že i neterčový podnět vyžaduje procesy
navazující na základní senzorickou diskriminaci, jejímž korelátem jsou již známé časné ERP
komponenty. Přinejmenším by se mohlo jednat o porovnání s informací uloženou v paměti,
konkrétně, že s neterčovým podnětem není spojena žádná akce a z toho vyplývající rozhodnutí
nic nedělat.
K ověření výše uvedeného předpokladu jsme zvolili opačný postup, než při identifikaci
vlny P3-like. Hledali jsme, ve vyšetřených strukturách místa, kde se odpovědi po terčovém a
neterčovém podnětu shodují. Pro zvýšení věrohodnosti analýzy jsme odpovědi po neterčových
podnětech rozdělili do dvou skupin. Do jedné skupiny jsme vybrali odpovědi na každý první
neterčový podnět prezentovaný po podnětu terčovém, označili jsme je jako „nehabituované“.
Do druhé skupiny jsme zařadili odpovědi na každý šestý, sedmý případně další neterčový
podnět, které následovaly po terčovém podnětu, označeny byly jako „habituované“ (příloha 3).
36
Příloha 3. Kukleta M, Brazdil A, Roman R, Jurak P. Identical event-related potentials to target
and frequent stimuli of visual oddball task recorded by intracerebral electrodes. Clinical
Neurophysiology 2003;114:1292-1297.
Do této studie byla zařazena data ze zrakového oddball úkolu od 20 pacientů (6 žen a
14 mužů). EEG signál jsme hodnotili v jeho celé šíři frekvenčního pásma 0,1-40 Hz. Celkem
bylo vyšetřeno 660 míst ve frontálních, temporálních a parietálních lalocích. Počet vybraných
EEG odpovědí po terčových podnětech byl průměrně 52. Do skupiny neterčových
nehabituovaných bylo vybráno průměrně 42 a habituovaných průměrně 89 odpovědí. Průběhy
zprůměrněných odpovědí po terčových a dvou skupinách neterčových podnětů byly vizuálně
porovnány a hodnoceny jako identické nebo neidentické dvěma nezávislými pozorovateli.
ERPs byly identifikovány na 530 místech. Identický průběh ERP po terčových a
neterčových podnětech byl pozorován v 88 případech, nejčastěji v amygdale, gyrus
parahipokampalis, gyrus temporalis superior, medius a inferior, gyrus fusiformis a lingualis,
méně často pak v senzorimotorickém kortexu, bazálním a dorzolaterálním prefrontálním
kortexu a hipokampu. Ve zbylých 442 případech se evokované odpovědi po obou typech
podnětů začaly ve svém průběhu lišit cca 300-470 ms od prezentace podnětů. Identické ERP
nebyly nalezeny v bazálních gangliích a parietálním kortexu.
Výsledky této studie jako první ukazují, že v určitých strukturách mozku probíhá
kognitivní zpracování senzorického podnětu bez vztahu k jeho významu (tzn. nezávisle
na způsobu odpovědi). I přes zdánlivou „nevýznamnost“ neterčového podnětu v tomto úkolu
se po jeho detekci velmi pravděpodobně musí aktivovat paměťové, případně rozhodovací nebo
kontrolní procesy, které jsou potřebné pro správné plnění úkolu. V porovnání s výskytem
odlišných ERPs se po obou typech podnětů shodné ERPs vyskytují 1) mnohem méně často a
2) hlavně v meziotemporálních strukturách. Uvedené skutečnosti by potom mohly vysvětlovat,
proč tyto elektrofyziologické koreláty nejsou pozorovatelné na skalpu.
III.4. Terčové podněty častěji indukují pokles výkonu EEG aktivity v alfa pásmu,
neterčové podněty naopak častěji nárůst výkonu EEG aktivity v alfa i beta pásmu
EEG aktivita během volního pohybu, paměťových a dalších kognitivních procesů
souvisí se změnami výkonu alfa, beta a theta oscilací. Změny výkonu v theta pásmu jsou
pozorovány v paměťových úkolech a souvisí pravděpodobně s kódováním a vybavováním
informací v pracovní paměti (Klimesch, 1999). Snížení výkonu v alfa pásmu (ERD;
37
„event-related desynchronization“) bývá u skalpových studií interpretováno jako
elektrofyziologický korelát aktivovaných procesů účastnících se zpracování senzorických a
kognitivních informací, pozornosti nebo spuštění motorické odpovědi. Nárůst výkonu
ve stejném pásmu (ERS; „event-related synchronization“) je chápán jako korelát inhibice
zpracovávání informací nebo paměťových procesů (Klimesch, 1996; Mazaheri a Picton, 2005).
Zvýšení výkonu v pásmu beta je pozorováno u řady senzorimotorických úkolů a v těchto
případech je spojováno s inhibicí kortikálních procesů. Někteří jiní autoři však fenomén ERS
v různých frekvenčních pásmech považují za korelát aktivního zpracování (Crone et al., 1998b,
2001).
Studium výkonových změn EEG signálu v průběhu oddball úkolu bylo v minulosti
realizováno výjimečně. Ve skalpových studiích při zrakové stimulaci byl pouze po terčových
podnětech prokázán signifikantní pokles výkonu v alfa pásmu (Sergeant et al., 1987),
u sluchové varianty úkolu nárůst výkonu v gama pásmu s maximem kolem 37 Hz (Gurtubay et
al., 2001). Jediná humánní intracerebrální studie, která se zaměřila na studium ERD/ERS
fenoménů v oddball úkolu, prokázala po terčové odpovědi v hipokampu ERS v pásmu theta a
ERD v pásmu alfa1 (Sochůrková et al., 2006). V této studii autoři analyzovali data pouze
6 pacientů a sledovali změny výkonu v definovaných frekvenčních pásmech a pouze
v meziotemporálních strukturách.
Některé změny výkonu ve frekvenčních pásmech s fixně nastavenými hranicemi však
nemusí dosáhnout hladiny signifikance. Toto omezení a malý počet vyšetřených pacientů
ve studii Sochůrková et al. (2006) byl pro nás podnětem k detailnější analýze tohoto fenoménu
(příloha 4).
Příloha 4. Roman R, Chládek J, Brázdil M, Jurák P, Rektor I, Kukleta M. Changes of
oscillatory activity in a visual oddball task (sEEG study). Homeostasis in Health and Disease
2006;44(4):169-171.
V tomto případě jsme data ze zrakového oddball úkolu analyzovali u 16 pacientů
(4 ženy a 12 mužů). Vyšetřili jsme celkem 150 míst ve frontálních, temporálních a parietálních
lalocích (66 v levé a 84 v pravé hemisféře). U každého pacienta jsme nejdříve provedli časově
frekvenční analýzu EEG odpovědi zvlášť pro terčové a neterčové podněty. Na jejím základě
jsme vybrali pro každého pacienta specifická, různě široká frekvenční pásma, ve kterých došlo
k signifikantní změně výkonu a která bylo možno zařadit do pásma alfa nebo beta. Cílem studie
38
byl popis struktur, ve kterých dochází k signifikantním změnám výkonu (ERD/ERS) a
porovnání změn výkonu mezi terčovou a neterčovou odpovědí.
Signifikantní změny výkonu EEG aktivity jsme pozorovali u všech pacientů po terčové
(95 míst) i neterčové odpovědi (84 míst) a to v různých mozkových strukturách v obou
hemisférách – gyrus cinguli, gyrus fusiformis, gyrus angularis, hipokampus, gyrus
parahipokampalis, gyrus temporalis superior, medius a inferior a frontoorbitálním kortexu.
Po terčových podnětech jsme v alfa pásmu pozorovali ERD v 59 % a ERS ve 41 % případů,
zatímco v beta pásmu ERD ve 46 % a ERS v 54 % případů. Po neterčových podnětech jsme
v alfa pásmu pozorovali ERD v 38 % a ERS v 62 % případů, v beta pásmu pak ERD ve 29 %
a ERS v 71 % případů. Po terčových podnětech v 17 případech a neterčových podnětech v 20
případech jsme na stejném kontaktu našli nárůst i pokles výkonu ve více pásmech současně.
Výsledky této studie ukázaly, že indukované EEG změny v oddball úkolu lze najít v celé
řadě struktur temporálního a frontálního laloku a rozšiřují tak nálezy předchozí studie
Sochůrkové a spol. (2006). Po terčových podnětech jsme častěji pozorovali ERD v alfa pásmu,
která by mohla odrážet procesy související s motorickou odpovědí. Na druhou stranu běžnější
nález ERS po neterčových podnětech by mohl v některých případech ukazovat na převahu
inhibičních procesů souvisejících s neprovedením stisku tlačítka. Zachycené změny výkonu
jsou zjevně vázány na odpověď subjektů, jejich jasná interpretace však doposud chybí. Z našich
výsledků také vyplývá, že sledování ERD/ERS v individuálně vybraných frekvenčních
pásmech může lépe zachytit i takové změny výkonu, které by nebyly patrné při fixně
nastavených hodnotách frekvenčních pásem.
III.5. Mapování „error“ potenciálů ukazuje účast řady kortikálních oblastí v mozkovém
systému detekce chybných odpovědí
Hodnocení správnosti vykonané odpovědi je pravděpodobně jedna z klíčových funkcí
lidského mozku. Na začátku 90. let byla popsána negativní komponenta skalpového ERP, která
se objevuje 50-100 ms po chybné motorické odpovědi (Falkenstein et al., 1991; Gehring et al.,
1993). Tato negativita je maximální frontocentrálně a označuje se termínem „error negativity“
(Ne) nebo „error-related negativity“ (ERN). Předpokládá se, že tento na chybu vázaný potenciál
je elektrofyziologickým korelátem detekce chybné odpovědi nebo kontroly odpovědi jako
takové (Coles et al., 2001; Falkenstein et al., 2000; Vidal et al., 2000). Detekci chyby někteří
autoři vysvětlují jako detekci neshody (konflikt) mezi provedenou chybnou odpovědí a neurální
reprezentací správné odpovědi vytvořené v průběhu kontinuálního zpracování podnětu
39
(Botvinick et al., 2001; Yeung et al., 2004). Lokalizační analýzy komponent skalpových ERP
ukazují na možné zdroje ERN v předním cingulárním kortexu nebo pre-suplementární
motorické oblasti (Holroyd et al., 1998; Luu et al., 2000; Ridderinkhof et al., 2004). Aktivaci
ACC po chybných odpovědích prokazují také některé fMRI studie (např. Braver et al., 2001;
Carter et al., 1998; Fiehler et al., 2004; Garavan et al., 2003). Méně často se aktivace objevuje
v pre-SMA, inzulárním kortexu, zadním cingulárním kortexu, thalamu, lobulus parietalis
inferior a gyrus supramarginalis, (Menon et al., 2001; Ullsperger a von Cramon, 2001).
Intracerebrální registrace ERP u epileptických pacientů během předoperačního videoEEG
monitorování nabízí možnost přímé lokalizace zdrojů evokovaných potenciálů.
V souvislosti s mapováním mozkových oblastí účastnících se zpracování chybných odpovědí
je bezesporu zajímavý výsledek naší intracerebrální studie z roku 2002. V průběhu zrakového
oddball úkolu totiž někteří pacienti také někdy chybně stiskli tlačítko po terčových podnětech.
Vybrali jsme 7 pacientů s dostatečným počtem chybných odpovědí (průměrně 9,3).
Zprůměrněním těchto chybných odpovědí jsme získali typické „error“ potenciály, tedy ERN.
Jejich zdroje jsme prokázali v řadě mozkových oblastí, zejména v gyrus cinguli anterior,
meziotemporálních a prefrontálních korových oblastech (Brázdil et al., 2002). Naše výsledky
však z důvodů použitého úkolu nebylo možno jednoduše srovnávat s ostatními ERP nebo fMRI
studiemi.
K odstranění této nevýhody jsme pro mapování intracerebrálních generátorů ERN
v nové studii (příloha 5) zvolili Go/NoGo úkol, používaný běžně v této oblasti výzkumu.
Příloha 5. Brazdil M, Roman R, Daniel P, Rektor I. Intracerebral error-related negativity in a
simple Go/NoGo task. Journal of Psychophysiology 2005;19:244-255.
V Go/NoGo úkolu byly na monitoru prezentovány střídavě dva typy podnětů - velké
písmeno „X“ a „K“ v poměru 4:1 (celkem 310 podnětů); přičemž „K“ nikdy dvakrát po sobě;
trvání podnětu 240 ms, interstimulus interval byl náhodně 650 ms, 1650 ms nebo 2650 ms.
Úkolem pacientů bylo stisknout co nejrychleji tlačítko pouze po písmenu „X“. Studie se
zúčastnilo 7 pacientů (1 žena a 6 mužů). Celkem jsme vyšetřili 574 míst frontálního,
temporálního a parietálního laloku. Při vlastní analýze jsme EEG data nejdříve filtrovali
(pásmová propust 2-10 Hz). EEG epochy délky 600 ms (300 ms před a 300 ms po stisku
tlačítka) jsme zprůměrnili od stisku tlačítka zvlášť pro správné odpovědi (tzn. stisk po „X“) a
pro chybné odpovědi (stisk po „K“) a vzájemně porovnali.
40
Průměrná chybovost našich pacientů byla 46±18,7 %. Generátory „error“ potenciálů
jsme nalezli v mnoha kortikálních oblastech – shodně s nálezy fMRI studií jsme je nejčastěji
identifikovali mediálně frontálně (gyrus cinguli anterior, pre-SMA, gyrus frontalis medialis),
dále pak v dorzolaterálním prefrontálním kortexu, orbitofrontálním kortexu, laterálním
temporálním kortexu a v izolovaném případě také v gyrus supramarginalis. Zajímavý byl nález
dvou zdrojů ERN v ACC – jeden v rostrální a druhý v kaudální části. To by mohlo souviset
s odlišnými procesy, tj. emočním zpracováním a specifickou detekcí chybné odpovědi, protože
rostrální část ACC funkčně souvisí s emocemi a dorzální část s kognitivními funkcemi a
řízením motoriky (Bush et al., 2000; Devinsky et al., 1995; Paus et al., 1993). Přítomnost ERN
v laterálním prefrontálním kortexu jsme prokázali i v naši předchozí studii (Brázdil et al., 2002).
Současná aktivace laterálního prefrontálního kortexu s ACC během chybných odpovědí byla
již prokázána Gehringem a Knightem (2000), ale její funkční význam není jasný.
Aktivace pre-SMA při kontrole provedení odpovědi je v fMRI studiích většinou
asociována se zpracováním konfliktu a inhibicí odpovědi (Braver et al., 2001; Menon et al.,
2001). Náš nález ERN v této oblasti však shodně se studií Garavana et al. (2004) potvrzuje její
účast i při detekci chyby. Na rozdíl od neurozobrazovacích studií se nám podařilo prokázat
ERN i ve frontoorbitálním kortexu, který je funkčně spojován s modulací impulzivity (Bechara
et al., 2000). Přítomnost „error“ potenciálů v řadě kortikálních oblastí je důkazem existence
rozsáhlé neuronální sítě, která může být v tomto případě podkladem mozkového systému
kontroly chybných odpovědí.
III.6. Po provedení správné odpovědi v rámci oddball úkolu se aktivují dvě funkčně
odlišné neuronální sítě, některé ze zúčastněných struktur jsou současně využívány i při
hodnocení odpovědí chybných
Při studiu neuronálních systémů, které se podílejí na hodnocení správnosti vykonané
odpovědi v jednoduchých kognitivních úkolech, se pozornost vědecké komunity většinou
zaměřuje na tzv. „error“ potenciály generované po chybné odpovědi. Jejich detailnější popis je
uveden v předchozím komentáři k práci Brázdila a spol. (2005) výše uvedené v III.5.
Elektrofyziologické koreláty vyvolané po provedení správné odpovědi jsou studovány
v mnohem menší míře. Ve skalpových studiích se popisuje negativní potenciál vyvolaný
po správně provedené odpovědi, tzv. „correct response-related negativity (Nc/CRN)“
s frontální (Mathalon et al., 2002; Meckler et al., 2011) a frontocentrální distribucí (Falkenstein
41
et al., 2000; Hajcak et al., 2005). Na základě těchto výsledků se za zdroj Nc považuje rostrální
cingulární kortex (Carter et al., 1998; Roger et al., 2010). Podobně jako u skalpových „error“
potenciálů lze u správných odpovědí po Nc nalézt i pozdější pozitivní komponentu, tzv. „correct
response positivity (Pc)“, která je studována minimálně a slouží obecně jako „klidová“ aktivita
při porovnávání s Pe (Bates et al., 2004). Na rozdíl od „error“ potenciálů, které jsou studovány
i s pomocí intrakraniálního EEG, analogické elektrofyziologické studie a hlubší interpretace
evokovaných potenciálů po správné odpovědi chybí.
K doplnění chybějících poznatků z oblasti hodnocení správnosti provedených odpovědí
při provádění kognitivních úkolů jsme opět analyzovali data pacientů vyšetřených v oddball
úkolu (příloha 6). Pro účely této studie jsme navíc využili skutečnosti, že kromě naprosté
většiny správných odpovědí pacienti také občas chybovali - po terčovém podnětu spínač
nestiskli nebo po častém podnětu spínač stiskli. Tři pacienti dokonce chybovali po terčových i
neterčových podnětech v dostatečném počtu vhodném pro zpracování ERP. Vedle mapování
oblastí aktivovaných po správných motorických reakcích po terčových odpovědích jsme tak
v této studii mohli současně posuzovat i možnou účast těchto oblastí v reakci na odpovědi
chybné, což doposud v intrakraniálních studiích popsáno také nebylo.
Příloha 6. Damborska A, Roman R, Brazdil M, Rektor I, Kukleta M. Post-movement
processing in visual oddball task - Evidence from intracerebral recording. Clinical
Neurophysiology 2016;127(2):1297-1306.
Ve studii byla analyzována data 18 pacientů (4 ženy a 14 mužů). Bylo vyšetřeno celkem
205 míst ve frontálních, temporálních a parietálních lalocích. EEG signál byl filtrován
ve frekvenčním pásmu 0,1-40 Hz. EEG epochy v trvání 1800 ms byly zprůměrněny od
okamžiku prezentace podnětu pro správné odpovědi po terčových podnětech a od stisku tlačítka
zvlášť pro odpovědi správné i chybné po obou typech podnětů. Počet chybných odpovědí
v dostatečném počtu pro zprůměrnění (min. 13 chyb) měli pouze 3 pacienti, u kterých bylo
možné srovnávat elektrofyziologické odpovědi po správných i chybných odpovědích.
Průměrná reakční doba u správných odpovědí po terčových podnětech v souboru 18
pacientů se pohybovala v rozmezí od 457±34ms po 644±78ms. Po zprůměrnění EEG epoch
těchto odpovědí od okamžiku stisku tlačítka byly identifikovány „post-movement“ potenciály
(= po stisku tlačítka, tj. po pohybu) na 131 místech mozku. Jejich generátory byly často
identifikovány v amygdale, hipokampu, laterálním temporálním kortexu, v dorsolaterálním
prefrontálním, předním cingulárním a orbitofrontálním kortexu. Morfologicky se nejčastěji
42
jednalo o monofazické (80%), méně často o bifazické potenciály (v naší studii jsme tyto
potenciály neoznačili zkratkami Nc/CRN nebo Pe pro jejich nejasný vztah k analogickým
skalpovým ERP). Průměrná latence vrcholů potenciálů byla 295±184 ms po pohybu. Ve 27 %
případů jim nepředcházely žádné ERP vlny v časovém intervalu od podnětu po stisk tlačítka.
Na 73 místech, kde byly identifikovány potenciály po pohybu u správných odpovědí
na terčové podněty, byly při zprůměrnění od podnětu nalezeny potenciály také po správné
odpovědi na neterčový podnět s latencí vrcholů (757±168ms) přesahující průměrnou reakční
dobu subjektu. Při porovnání s latencemi potenciálů po terčových podnětech měřených
při stejném způsobu zprůměrnění se ukázalo, že pozdní potenciály po neterčových podnětech
vykazovaly bez specifické vazby na struktury kratší latenci o 174±118ms ve 34 místech, delší
latenci o 166±88ms ve 26 místech a přibližně stejnou latenci v 7 místech.
Reakční doba po chybných odpovědích u třech pacientů byla delší v porovnání s reakční
dobou po odpovědích správných. Pouze u dvou pacientů byly u chybných odpovědí („incorrect
rejection, false alarms“) nalezeny pozdní potenciály a to v předním cingulárním,
dorsolaterálním prefrontálním, primárním motorickém a laterálním temporálním kortexu.
V těchto místech byly nalezeny současně i „post-movement“ potenciály po správné odpovědi
na terčový podnět. U jednoho pacienta byly latence vrcholů těchto potenciálů delší než
u potenciálů po správných odpovědích, u druhého pacienta se latence vrcholů nelišily více než
o 20 ms.
V této studii se podařilo prokázat, že integrální součástí neuronálních sítí aktivovaných
po správném provedení odpovědi v rámci oddball úkolu je kromě již dříve ve skalpových
studiích identifikovaného předního cingulárního kortexu (Roger at al., 2010) také řada dalších
struktur zejména temporálního a frontálního laloku. Účast některých z nich byla specifická
pouze pro terčové odpovědi, u jiných byla na typu podnětu nezávislá. Z tohoto lze usuzovat
na existenci dvou funkčně odlišných neuronálních sítí, které částečně sdílejí některé anatomické
struktury. Aktivace neuronální sítě v prvním případě pak může odpovídat za mentální počítání
podnětů s nezbytnou aktualizací pracovní paměti, případně zpracování somatosenzorických
aferentních informací z aktivovaných svalů. Elektrofyziologické koreláty na typu podnětu
nezávislé pak mohou reprezentovat procesy hodnotící správnost provedené odpovědi
v porovnání s informací o významech jednotlivých podnětů uložených v paměti. Nález
mozkových oblastí, které jsou aktivovány jak u správných tak u chybných odpovědí, potvrzuje
dřívější předpoklady ze skalpových studií, že generátory „error“ a „correct“ potenciálů se
pravděpodobně překrývají (Wessel et al., 2012).
43
III.7. Přední cingulární kortex hraje významnou roli v komunikaci s temporálními
oblastmi mozku
V současné době se předpokládá, že podkladem kognitivních funkcí jsou mechanismy
funkční integrace četných neuronálních populací aktivovaných v průběhu zpracování
informací. Otázkou zůstává, zda v integraci neuronálních sítí velkého rozsahu („large-scale
cognitive networks“) hraje klíčovou roli nějaká struktura nebo je to vlastnost kognitivních sítí
jako takových (Bressler a Kelso, 2001; Fries et al., 2001; Jensen et al., 2007; Varela et al.,
2001). Jedním z kandidátů strukturální lokalizace integrativních funkcí je přední cingulární
kortex (ACC). Anatomicky je propojen s celou řadou oblastí mozku a jeho aktivaci v průběhu
kognitivních úkolů ukazuje řada studií (např. Allman et al., 2001; Paus, 2001; van Veen a
Carter, 2002). Experimentální doklady potvrzují jeho funkční účast např. při řešení problémů,
rozpoznání chyb, rozpoznání a řešení konfliktních situací nebo při adaptivní odpovědi
v měnících se podmínkách (Allman et al., 2001; Posner et al., 2007). Ze studií vyplývá, že ACC
se podílí na motorické, kognitivní a motivační složce chování. Paus (2001) uvádí, že právě účast
na třech složkách chování odlišuje ACC od ostatních frontálních kortikálních oblastí a
umožňuje tak této struktuře převádět záměry k akcím. Dle Rueda a spol. (2004) spočívá hlavní
příspěvek ACC k mozkovým funkcím v jeho řízení informačních vstupů z prostředí a tím
k omezení vzniku konfliktních behaviorálních odpovědí.
Cingulární kortex, zejména jeho přední část, byl vedle dalších frontálních korových
oblastí často vyšetřován v rámci hledání epileptického ložiska u naší skupiny epileptických
pacientů, kteří prováděli zrakový oddball úkol. Současná představa o klíčovém postavení ACC
v integraci aktivit různých neuronálních populací nás vedly k myšlence ověřit hypotézu, že
míra funkční interakce ACC se vzdálenými mozkovými oblastmi odlišuje tuto strukturu
od ostatních frontálních oblastí (příloha 7). Za tímto účelem jsme porovnávali úroveň
synchronizace intracerebrálního EEG signálu ve frekvenčním pásmu beta2 (pásmová propust
25-35 Hz) mezi frontálními a temporálními laloky, přičemž některé frontální kontakty se
nacházely v ACC a další v jiných frontálních oblastech. Frekvenční pásmo beta2 jsme vybrali
na základě nálezu vyššího zastoupení synchronních oscilací právě v tomto pásmu v naší
předchozí studii zaměřené na analýzu synchronizací mezi frontálními a temporálními oblastmi
v různých frekvenčních pásmech (Kukleta et al., 2009).
44
Příloha 7. Kukleta M, Bob P, Brazdil M, Roman R, Rektor I. The level of frontal-temporal
beta-2 band EEG synchronization distinguishes anterior cingulate cortex from other frontal
regions. Consciousness and Cognition 2010;19:879-886.
V předkládané studii jsme vyšetřili EEG data ze zrakového oddball úkolu u 8 pacientů
původního souboru. V prvním kroku jsme vybrali 180 fronto-temporálních párů
intracerebrálních kontaktů. EEG signál ve frontálních lalocích byl ve 41 případech snímán
v předním cingulárním kortexu, v 60 případech v orbitofrontálním kortexu, v 41 případech
v dorzolaterálním prefrontálním kortexu a v 38 případech v mesiálním frontálním kortexu.
Druhý kontakt vybraných párů se nacházel v temporálním laloku (většinou v laterálních
korových oblastech, méně často v gyrus parahipokampalis a fusiformis, amygdale a
hipokampu). Následoval výpočet korelačních koeficientů filtrovaného (25-35 Hz) EEG signálu
a porovnání jejich průměrných hodnot ve všech 180 párech ve třech funkčně odlišných
obdobích: a) v osmnácti 30s úsecích vybraných v průběhu celého oddball úkolu nezávisle na
okamžiku prezentace podnětů; b) v 1s úseku předcházejícím prezentaci podnětu; c) v 1s úseku
po prezentaci podnětu. V případě b) a c) jsme výpočet korelace provedli na zprůměrněných
EEG odpovědích a to jenom po neterčových podnětech (nebyly kontaminovány motorickou
aktivitou).
Výsledky studie ukázaly, že průměrné hodnoty korelačních koeficientů ve frontotemporálních
párech, které měly frontální kontakt v ACC, byly signifikantně vyšší než v párech
s kontaktem v jiných frontálních oblastech a to ve všech třech funkčně odlišných obdobích,
tedy v průběhu celého experimentu, v období zvýšeného očekávání dalšího podnětu a v období
senzorického a kognitivního zpracovávání neterčového podnětu. V souladu s výsledky jiných
autorů (Paus, 2001; Posner et al., 2007) tyto nálezy dokládají významnou roli předního
cingulárního kortexu v neuronálních sítích velkého rozsahu, která může být podkladem
integračních funkcí v průběhu kognitivních procesů.
III.8. Při zpracování podnětů spojených s aktivní odpovědí se v rámci frontálního laloku
informace šíří směrem z gyrus cinguli do dorzolaterálního prefrontálního kortexu
Jak bylo uvedeno dříve, podněty v oddball úkolu spouští v mozku řadu procesů, které
jsou podkladem časného senzorického zpracování, pozornosti a exekutivních funkcí. Přední
cingulární kortex (ACC) a dorzolaterální prefrontální kortex (DLPFC) jsou považovány
45
za struktury účastnící se pozornostních procesů vyvolaných neočekávanými změnami
v prostředí nebo procesů souvisejících s výběrem správné odpovědi. V těchto oblastech byl
opakovaně prokázán generátor hluboké vlny P3a (Baudena et al., 1995; Brázdil et al., 1999;
Halgren et al., 1998). Předpokládá se, že tyto dvě oblasti při řízení chování spolupracují,
přičemž ACC je důležitý pro sledování zpracování podnětu a s tím související vhodné odpovědi
(Shallice, 1988; Shallice et al., 1989). V případě potřeby ACC zřejmě aktivuje DLPFC
(a obecně arousal systém), jehož prostřednictvím by se selektivně posílily procesy související
se zpracováním podnětu a jeho kortikální reprezentací v zadních kortikálních oblastech, aby
mohly být splněny vnitřní cíle nebo dohodnuté instrukce (Barch et al., 2001; Botvinick et al.,
1999; Carter et al., 1998; Cohen et al., 2000). V tomto smyslu může být ACC chápáno jako
klíčová oblast při časné alokaci pozornosti (Gitelman et al., 1999; Woldorf et al., 2001).
Hierarchické funkční uspořádání ACC a DLPFC v kognitivních úlohách bylo
prokázáno i v rámci oddball úkolu s využitím fMRI (Brázdil et al., 2007). Nízké časové
rozlišení této techniky však limituje prokazování směru informačního toku (efektivní funkční
konektivity) mezi kortikálními oblastmi během časově krátkého zpracování podnětů. Pro jeho
detailnější studium lze využít výpočet směrové transformační funkce, anglicky „directed
transformation function“ (DTF) z EEG záznamů (Kaminski a Blinowska, 1991). S použitím
tohoto výpočetního algoritmu bylo např. prokázáno, že tok informací z parietookcipitálních
do frontálních oblastí je převažující v období před usnutím nebo při kódování zrakových
informací, zatímco opačný směr šíření informací je typický po usnutí (De Gennaro et al., 2004;
Babiloni et al., 2006).
Relativně častá současná explorace předního cingulárního kortexu a dorzolaterálního
prefrontálního kortexu u naší skupiny epileptických pacientů opět exkluzivně umožnila využít
intracerebrálně registrovaný EEG signál tentokrát pro studium efektivní funkční konektivity
mezi těmito dvěma oblastmi během zpracování podnětů v průběhu oddball úkolu (příloha 8).
S použitím DTF algoritmu jsme se snažili ověřit hypotézu, že ACC ovlivňuje oscilační
aktivitu DLPFC.
Příloha 8. Brazdil M, Babiloni C, Roman R, Daniel P, Bares M, Rektor I, Eusebi F, Rossini
PM, Vecchio F. Directional Functional Coupling of Cerebral Rhythms Between Anterior
Cingulate and Dorsolateral Prefrontal Areas During Rare Stimuli: A Directed Transfer Function
Analysis of Human Depth EEG Signal. Human Brain Mapping 2009;30:138-146.
46
Za tímto účelem jsme analyzovali data ze zrakového oddball úkolu 8 pacientů (2 ženy
a 6 mužů), u kterých byly v rámci předoperační léčby epilepsie současně vyšetřeny obě oblasti.
Do analýzy byla použita EEG data po terčových i neterčových podnětech pouze z ACC
(area 32 a 24) a DLPFC (area 9, 44, 45 a 46) a to v jednom časovém okně 500 ms
před okamžikem prezentace podnětu a ve druhém časovém okně 500 ms po prezentaci podnětu.
Dle standardizovaného postupu (Kaminsky a Blinowska,1991) byla EEG data nejdříve
normalizována a poté použita jako vstupní parametry vícerozměrného autoregresního modelu
pro výpočet DTF. Jednoduše lze říci, že použitý model určuje informační tok mezi místy A a B
prostřednictvím výpočtu pravděpodobnosti, se kterou lze průběh EEG dat v místě A predikovat
na základě průběhu EEG dat v místě B a naopak. DTF výpočet byl proveden pro terčové a
neterčové EEG odpovědi ve frekvenčním pásmu theta, alfa, beta a gama.
V časovém okně 500 ms před prezentací podnětů byly DTF hodnoty pro směr toku
informací ACC→DLPFC a DLPFC→ACC ve všech frekvenčních pásmech po terčovém i
neterčovém podnětu přibližně shodné. V časovém okně 500 ms po prezentaci podnětů se DTF
hodnoty zvýšily pro směr toku informací ACC→DLPFC, ale jen po terčovém podnětu, přičemž
signifikantní změny bylo dosaženo pouze v pásmu theta.
Výsledky naší studie tedy potvrzují hypotézu, že při odpovědi mozku na terčový podnět
v průběhu oddball úkolu, přední cingulární kortex ovlivňuje oscilační aktivitu dorzolaterálního
prefrontálního kortexu zejména v theta frekvenčním pásmu. Na událost vázané modulace právě
tohoto theta rytmu jsou důležitým elektrofyziologickým korelátem mozkových procesů
souvisejících s integrací senzorimotorických a kognitivních informací (Basar et al., 1999b;
Brankack et al., 1996; Pfurtscheller a Lopes da Silva, 1999). Předpokládá se, že neuronální
oscilace v theta pásmu souvisejí se zaměřením pozornosti, pracovní pamětí, kódováním
epizodické paměti a řízením akcí (Gevins a Smith, 2000; Sauseng et al., 2002, 2004;
Sochůrková et al., 2006). V těchto souvislostech pak v naší studii identifikovaný informační
tok z ACC do DLPFC může souviset s pozorností, aktualizací pracovní paměti a řízením
motorické odpovědi na obtížně předvídatelný vzácný (terčový) podnět.
47
IV. Závěr
Intracerebrální elektroencefalografie je vyšetřovací metoda, která je používána
u pacientů s neurologickým onemocněním výhradně na základě přesně definovaných
klinických požadavků. Přesto tato metoda nabízí jedinečnou příležitost studovat elektrickou
aktivitu neuronálních populací mozku u člověka při vědomí. V neurovědeckém výzkumu tvoří
sice minoritní podíl, poskytuje však unikátní pohled na mechanismy mozkových funkcí, které
nelze studovat u zvířat. Přispívá tak zásadně k detailnějšímu pochopení fungování lidského
mozku, což je základním předpokladem pro vývoj sofistikovanějších diagnostických a
terapeutických postupů použitelných při jeho poruchách.
Při zpracování intracerebrálních EEG dat snímaných v průběhu kognitivních úkolů se
používají některé metodické postupy běžně využívané ve skalpových EEG studiích.
Zprůměrnění několika desítek EEG odpovědí od okamžiku prezentace podnětu umožňuje např.
analyzovat evokované (ERP) a indukované (ERD/ERS) děje a určit jejich intracerebrální
distribuci. Pro svoje specifické vlastnosti, jako je např. vysoký poměr signál-šum, je možné
iEEG data analyzovat i způsobem, který u skalpového EEG signálu lze použít výjimečně, pokud
vůbec. Příkladem mohou být námi navržené analýzy pomalých kognitivních potenciálů
v podskupinách s malým počtem EEG odpovědí. Nevýhodou intracerebrálních EEG dat je
nemožnost jejich matematického zprůměrnění ve skupině pacientů, protože se hluboké
elektrody implantují do různých oblastí a nikdy tak nejsou vyšetřena naprosto identická místa.
Pro identifikaci zákonitých jevů je potom nutné např. vizuální porovnání EEG odpovědí
snímaných z různých oblastí mozku u jednotlivých pacientů a hledání takových
elektrofyziologických projevů, jako je např. morfologie a latence ERP komponent, které jsou
vyšetřeným oblastem a skupině pacientů společné. Všechny zmiňované postupy jsme použili
v naší práci s cílem hlouběji pochopit vztah mezi elektrofyziologickými koreláty
registrovanými intracerebrálně a předpokládanými mentálními operacemi, přičemž část našich
nálezů je komentována v předloženém souboru prací.
Při výzkumném použití iEEG platí dvojnásob, že vhodně navržený úkol a perfektně
dodržená pravidla získávání dat s detailní dokumentací okolností nahrávání mohou poskytnout
data s vysokým informačním obsahem. Je potom jen otázkou času a vůle nové zákonitosti
objevovat a popsat.
48
Příloha č. 1
Roman R, Brazdil M, Jurak P, Rektor I, Kukleta M. Intracerebral P3-like waveforms and the
length of the stimulus-response interval in a visual oddball paradigm. Clinical Neurophysiology
2005;116:160-171. IF (2005) = 2,640
Intracerebral P3-like waveforms and the length of the stimulus–response
interval in a visual oddball paradigm
Robert Romana,*, Milan Bra´zdilb
, Pavel Jura´kc
, Ivan Rektorb
, Miloslav Kukletaa
a
Department of Physiology, Medical Faculty, Masaryk University, Komenske´ho na´m. 2 662 43 Brno, Czech Republic
b
Department of Neurology, St Anne Hospital, Masaryk University, Brno, Czech Republic
c
Institute of Scientific Instruments, Academy of Sciences, Brno, Czech Republic
Accepted 15 July 2004
Available online 1 September 2004
Abstract
Objective: This study investigated the possible linkage of intracerebrally recorded P3-like waveforms to the processes induced by
stimulus perception or motor response formation.
Methods: Event-related potentials were recorded from 560 cerebral sites in 17 patients suffering from intractable epilepsy during visual
oddball task. Potentials evoked by the target stimuli were sorted according to button-pressing response times, and the P3 waveform was
analyzed both in stimulus-locked and response-locked averages, which were separately averaged for fast and slow responses.
Results: P3-like waveforms were identified in 180 sites in 17 patients. Three different types of P3-like waveforms, diffusely distributed
within the brain, were found: (1) time-locked to the stimulus (30 sites in 11 patients); (2) time-locked to the motor response (52 sites in
13 patients); and (3) with ambiguous time relationship to stimulus and motor response (98 sites in 16 patients).
Conclusions: The intracerebral P3-like waveform could represent different processes involved in performing active oddball tasks.
Therefore, our results support the hypothesis that the P3 waveform registered by surface electrodes could be a heterogeneous phenomenon.
Significance: These results provide evidence that the P3 waveform is not only related to stimulus processing, which differs from what has
been generally claimed in the literature.
q 2004 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: P3 waveform; ERPs; Intracerebral recordings; Visual oddball paradigm
1. Introduction
Event-related potentials (ERPs) are composed of a series
of waveforms, each defined by its latency, polarity, scalp
topography, and behavioural correlates (Halgren et al.,
1998). Successive waveforms can be divided into early
sensory, or exogenous, potentials primarily influenced by
physical characteristics of the eliciting stimulus (Cracco and
Cracco, 1976; Jewett et al., 1970), and late cognitive, or
endogenous, potentials, primarily influenced by cognitive
processes (Donchin et al., 1978; Halgren et al., 1998;
Hillyard et al., 1978; Sutton et al., 1965).
One of the most studied waveforms of endogenous or
‘cognitive’ ERPs is the P3 (P300) waveform, which was first
described by Sutton and colleagues in 1965. It is usually
elicited in the oddball paradigm, wherein two stimuli are
presented in a random order with different probabilities. In
the active form of this paradigm, a subject is required to
discriminate between the infrequent target stimulus and the
frequent standard stimulus by noting the occurrence of the
target, usually by pressing a button or mentally counting.
The P3 waveform can also be obtained with alternative
tasks: a passive oddball paradigm, a single stimulus
paradigm, or a 3-stimulus paradigm (Bennington and
Polich, 1999; Jeon and Polich, 2001; Polich, 1998). Using
the 3-stimulus paradigm, two types of P3 waveform have
been described: the novelty ‘P3a’ waveform, with frontal/
central maximum scalp distribution, elicited by novel
Clinical Neurophysiology 116 (2005) 160–171
www.elsevier.com/locate/clinph
1388-2457/$30.00 q 2004 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.clinph.2004.07.016
* Corresponding author. Tel.: C420-542-496-818; fax: C420-542-126-
651.
E-mail address: roman@med.muni.cz (R. Roman).
stimuli (Squires et al., 1975), and the later ‘P3b’ waveform,
with the largest amplitudes over parietal areas, elicited by
the infrequent target stimuli (Comerchero and Polich,
1999). A kind of P3 potential has even been recorded in a
modified visual oddball paradigm after subliminally presented
target stimuli (Bra´zdil et al., 2001).
In a task requiring specific motor response to a target
stimulus, distinct movement-related ‘P3-like’ potentials
(MRPs) are also generated intracerebrally that coincide
with P3 potentials (Bra´zdil et al., 2003). The temporal
occurrence of these MRPs is related to the occurrence of the
movement and, therefore, they are time-locked to the
button-pressing (Vaughan et al., 1968). The P3 usually
coincides with the movement or precedes it, and thus these
MRPs can distort the P3 scalp topography and can modify
P3 amplitude and latency (Bra´zdil et al., 2003; Jentzsch and
Sommer, 2001; Kok, 1988; Salisbury et al., 2001). Since the
scalp potentials only reflect the sum of electrical activity
propagated in a given area, it is hard to distinguish this MRP
from the cognitive P3 waveform. This MRP contamination,
however, can be at least partially eradicated by the use of a
trial-by-trial reaction time (RT) matching procedure, which
creates a variance-matched and temporally aligned MRPcorrection
waveform (Kok, 1988; Salisbury et al., 2001).
The P3 latency ranges from 250 to 600 ms after the
stimulus onset, and can vary depending on the task
complexity, stimulus, and subject (Comerchero and Polich,
1998; Mertens and Polich, 1997; Polich, 1998). The P3
latency is considered a measure of the stimulus classification
speed (Kutas et al., 1977; Polich, 1986a,b). It is
generally unrelated to response selection and execution
processes (McCarthy and Donchin, 1981; Pfefferbaum
et al., 1986), and is therefore, independent of behavioural
RT (Duncan-Johnson, 1981). This view is supported by
many studies where the manipulations with the stimuli
resulted in an increase of the RT as well as of the P3 latency:
the reduction of stimulus intensity (Verleger et al., 1991),
the reduction of stimulus discriminability (Novak et al.,
1990), increasing the number of irrelevant stimuli while
randomly changing the location of the relevant stimulus
(Magliero et al., 1984), presenting flanking stimuli to which
the alternative response is assigned (Smid et al., 1990). In
some other studies, the manipulation with responses was
used in order to increase the RT but constant P3 latency was
observed: using spatially incompatible stimulus–response
(SR) assignments (Fitzpatrick et al., 1988). However, when
presenting the stimulus on the side of the alternative
response (Nandrino and El Massioui, 1995) or when
increasing the number of response alternatives (Falkenstein
et al., 1994a), the P3 latency became longer. There is some
evidence that P3 latency is sensitive to response processing
time, particularly when the response times in a given task
are generally fast (Verleger, 1997). Indeed, the results of
simple RT tasks suggest that motor execution can
influence P3 latency. This effect, however, was observed
only when stimulus evaluation demands were minimal.
Increased stimulus evaluation demands are sufficient to
mask the motor execution effect on P3 latency. In general,
motor execution processes would not be considered a
confounding factor in most P3 studies (Doucet and
Stelmack, 1999; Rektor et al., 2003).
In studies using depth electrode recordings, P3 generators
were found in different brain structures (Baudena et
al., 1995; Bra´zdil et al., 1999; Halgren et al., 1995a,b). Two
characteristic, modality non-specific intracerebral patterns
of P3 were described. First, the triphasic waveform with
sharp negative, positive, and negative peaks termed N2aP3a-SW,
associated with the system for the orientation of
attention, was observed in the middle temporal, posterior
parahippocampal and fusiform gyrus sites, and in dorsolateral
prefrontal, inferior parietal and cingulate cortices.
Second, a broad P3b waveform, coupled to the eventencoding
system, was found in the hippocampus, amygdala,
superior temporal sulcus, lateral orbitofrontal cortex, and
intraparietal sulcus (Bra´zdil et al., 1999; Halgren et al.,
1995a,b). Due to their similarity to the scalp-recorded P3
waveforms, these were labelled as the depth P3a and depth
P3b, or generally P3-like waveforms. Their contribution to
the scalp-recorded P3s remains unclear.
To date, the P3 waveform is generally viewed as
reflecting decision or cognitive closure of the stimulus
identification. It has also been linked to orientation,
attentional mechanisms, and the update of working memory
(Paller et al., 1987; Squires et al., 1975). This relationship
between the P3 response and the fundamental aspects of
cognitive function is used in clinical practice for the
assessment of cognitive ability in normal subjects as well as
in patients with neurological and psychiatric diseases. To
obtain the most uniform P3 responses, specific stimulus
characteristics, task factors, subject variables, and proper
electrophysiological recording are recommended (Polich,
1998).
In the oddball paradigm, where recognition of the target
stimulus is indicated by pressing the button, two events well
defined in time are recorded simultaneously together with
EEG. They are the stimulus presentation, which starts
processing of this information and the button pressing which
stands for the signal of the subject that he/she recognized the
target stimulus. Approximately 50 target stimuli are used in
order to be randomly presented within the oddball task. The
behavioural RT, i.e. the time from the stimulus presentation
to the moment of the button pressing, can vary across trials
in each subject for 50–200 ms even though the subjects are
instructed to respond as fast as possible. Therefore we can
sort all 50 responses into fast and slow subgroups and
average them separately. If the studied waveform of ERP
represents process time locked to the stimulus, its peak
latency will be the same in fast and slow subgroups
averaged with the stimulus as a trigger. Back averaging,
with the button pressing as a trigger, would reveal, in this
case, that the peak latency of the waveform is longer in slow
averages and shorter in fast averages. If the waveform
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171 161
represents process time locked to the button pressing, the
peak latency will be longer in slow and shorter in fast
subgroups when averaged with the stimulus as a trigger. The
averaging with the button pressing as a trigger would then
reveal the same peak latency.
In our study, we analyzed intracerebrally-recorded
P3-like waveforms obtained in a standard visual oddball
paradigm in epileptic patients during their presurgical
diagnostic exploration. We were interested in whether the
unequivocal relationships between P3 latency and the
processes underlying the experimental task, found in scalp
recordings, could be also demonstrated on the depthrecorded
P3-like waveform. To answer this question, we
investigated the character of the relationship between the P3
latency and the SR interval. In each subject, we created one
averaged curve from trials with short SR intervals and
another curve from trials with long SR intervals. In both
cases, two averaging procedures were performed: with the
stimulus onset as a trigger and with button-pressing as a
trigger. The comparison of P3-like waveforms obtained
under these conditions enabled us to evaluate the possible
link between them and the processes induced by stimulus
perception or motor response formation.
2. Methods
2.1. Subjects
Seventeen patients (4 females and 13 males; see Table 1)
with medically intractable epilepsy participated in the study
(their mean age was 29.5G5.5 years). All the subjects had
normal or corrected-to-normal vision. Informed consent
was obtained from each subject prior to the experiment,
and the study received the approval of the Ethical
Committee of Masaryk University.
2.2. EEG recordings
A total of 89 depth electrodes were implanted to localize
the seizure origin prior to surgical treatment. 560 sites of the
frontal, parietal, and temporal lobes (248 in the right and
312 in the left hemisphere; see Table 1) were explored with
standard MicroDeep semiflexible multilead electrodes, each
with diameters of 0.8 mm. Recordings contacts were
2.0 mm in length, and successive contacts were separated
by 1.5 mm. Lesional anatomical structures and epileptogenic
zone structures were not included in the analysis. The
exact position of the electrodes and their contacts in the
brain were verified using postplacement MRI with electrodes
in situ.
The EEG signal was recorded simultaneously from
various intracerebral structures and from CPz scalp electrodes
(situated between Cz and Pz) during seizure-free periods
using the 64-channel Brain Quick EEG system (Micromed).
All the recordings were monopolar with respect to a
reference electrode on the right processus mastoideus and
were acquired at a 128 Hz sampling rate. All impedances
were less than 5 kU. Button-pressing was recorded on a
separate channel.
2.3. Behavioural task
A standard visual oddball paradigm was performed: two
types of stimuli -target and frequent- were randomly
presented in the centre of the screen. Clearly visible
uppercase yellow letters X (target) and O (frequent) were
presented on a white background as the experimental stimuli.
Table 1
Age, sex and handedness for 17 epileptic patients, along with the number of intracranial electrodes, recorded sites and brain lobes investigated
Patient ID Age (years) Sex Handedness Electrodes
left/right hemisphere
Recording sites
left/right hemisphere
Brain lobes
BS 25 M R 4/1 14/2 T0
, T, P0
DJ 20 M R 1/1 13/13 F0
, F, T0
, T
DM 28 M R 1/1 12/15 T0
, T
FP 37 M R 1/6 15/26 F0
, F, T0
, T
HZ 45 M R 5/4 23/16 F0
, F, T0
, T
HP 30 M R 5/2 28/12 F0
, F, T0
, T
JI 31 F R 0/6 0/41 F, T
KM 32 M R 1/1 14/14 F0
, F, T0
, T
KP 30 M R 0/4 0/40 T, P
KJ 30 M R 4/1 27/15 F0
, F, T0
, T
LL 27 F R 0/4 0/22 T
PM 25 M L 6/4 24/17 F0
, F, T0
, T
SM 34 M R 0/5 0/34 F, T, P
SO 30 F R 3/1 15/11 F, T0
, T, P0
SI 23 M R 4/5 16/25 F0
, F, T0
, T
VH 28 F R 7/0 41/0 F0
, T0
VP 27 M R 1/0 15/0 F0
, T0
F, frontal lobe; T, temporal lobe; P, parietal lobe; symbol (0
) indicates left hemisphere.
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171162
The duration of stimuli exposure was constant at 200 ms; the
ratio of target to frequent stimuli was 1:5. The interstimulus
interval varied randomly between 2 and 5 s. Each subject was
instructed to respond to the target stimulus as quickly and as
accurately as possible by pressing a microswitch button in the
dominant hand and, at the same time, he/she was instructed to
silently count the target stimuli.
2.4. Data analysis
ScopeWin software was used for further off-line signal
analysis. Each 1.7 s stimulus epoch beginning 500 ms prior
to the stimulus onset (used as a prestimulus baseline; in
figures only 300 ms before the stimulus onset is presented)
was inspected visually and further processing was performed
with artefact-free EEG periods. Because P3-like
waveforms have a relatively low frequency (Polich, 1998), a
digital filter with a bandpass of 0.5–5.5 Hz was used for the
offline analysis.
In each subject, approximately 50 trials were averaged
for target and 150 trials for frequent stimuli (unsorted
averages; sections A in Figs. 1–4). The identification of P3like
waveforms in the latency range of 250–600 ms was
based on a visual evaluation of the records by two
independent experts. To reduce the contamination of P3like
waveforms by movement-accompanying slow potentials
(Rektor et al., 1998), those with a latency longer than
the mean SR interval of each patient were excluded from the
analysis (11 from 191 cases). In each target trial, the SR
interval was measured. From all the target trials in each
patient, two subgroups with shorter and longer SR intervals
were created arbitrarily (see Table 2). In creating the
subgroups, we attempted to meet the following general
demands: sufficient number of trials for averaging in each
subgroup; identical number of trials in both subgroups;
corresponding range of SR intervals in each subgroup; and
the largest possible difference of the mean values of SR
intervals between the two subgroups. Selected trials were
then averaged in two procedures: (i) with the stimulus onset
as a trigger (sorted averages; sections B1 in Figs. 1–4),
(ii) with the button-pressing as a trigger (sorted averages;
sections B2 in Figs. 1–4). The peak P3 latencies in each
subgroup for both averaging procedures were then analyzed.
The differences of P3 peak latencies were considered as
identical if the difference between them reached no more
than 8 ms (with the sampling rate of 128 Hz, the difference
of 8 ms corresponds to the difference of 1 sample).
For statistical purposes, one-way ANOVA, post hoc
Spjotvoll/Stoline test (a variant of Tukey’s HSD test) and
regression analysis were performed. Statistics were obtained
using the routines included in the Statistica program
(StatSoft).
3. Results
The mean number of averaged target trials in each
subgroup of patients was 16G3 (minimum 9, maximum 24;
Fig. 1. Three examples of P3-like waveforms time-locked to the stimulus. Coronal sections representing electrode and contact position in each patient are not
proportional. Patient KJ (C0
2, left parahippocampal gyrus), patient SI (G1, right anterior cingulate gyrus) and patient VH (O0
5, left orbitofrontal cortex). (A)
ERPs to target (thick line) and non-target (thin line) stimulus, unsorted averages. (B1, B2) Target ERPs from the same sites and patients averaged from the
subgroup of trials with short SR intervals (thick line) and long SR intervals (thin line), with the stimulus onset as a trigger (on the left, B1) and the buttonpressing
as a trigger (on the right, B2). The peak P3 latency difference between the two subgroups averaged with the stimulus as a trigger was 0 ms (KJ), 0 ms
(SI), 8 ms (VH), whereas averaged with the button-pressing as a trigger it was 47 ms (KJ), 47 ms (SI) and 63 ms (VH). Symbol : indicates P3-like waveform.
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171 163
Fig. 3. Three examples of P3-like waveforms with ambiguous time relationship to stimulus and motor response. Coronal sections representing electrode and
contact position in each patient are not proportional. Patient HZ (B3, right hippocampus posterior), patient SO (X6, right putamen) and patient VH (O0
8, left
ventrolateral prefrontal cortex). (A) ERPs to target (thick line) and non-target (thin line) stimulus, unsorted averages. (B1, B2) Target ERPs from the same sites
and patients averaged from the subgroup of trials with short SR intervals (thick line) and long SR intervals (thin line), with the stimulus onset as a trigger (on the
left, B1) and the button-pressing as a trigger (on the right, B2). The peak P3 latency difference between the two subgroups averaged with the stimulus as a
trigger was 47 ms (HZ), 24 ms (SO), 31 ms (VH), whereas averaged with the button-pressing as a trigger it was 31 ms (HZ), 24 ms (SO) and 31 ms (VH).
Symbol : indicates P3-like waveform.
Fig. 2. Three examples of P3-like waveforms time-locked to the motor response. Coronal sections representing electrode and contact position in each patient
are not proportional. Patient SM (G3, right posterior cingulate gyrus), patient PM (O2, right rectus gyrus) and patient HP (F11, right premotor cortex,
Brodmann’s area 6). (A) ERPs to target (thick line) and non-target (thin line) stimulus, unsorted averages. (B1, B2) Target ERPs from the same sites and
patients averaged from the subgroup of trials with short SR intervals (thick line) and long SR intervals (thin line), with the stimulus onset as a trigger (on the
left, B1) and the button-pressing as a trigger (on the right, B2). The peak P3 latency difference between the two subgroups averaged with the stimulus as a
trigger was 47 ms (SM), 40 ms (PM) and 63 ms (HP), whereas averaged with the button-pressing as a trigger it was 0 ms (SM), 0 ms (PM), 8 ms (HP). Symbol
: indicates P3-like waveform.
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171164
see Table 2). We compared the morphology of P3-like
waveforms obtained from our subgroups with short and long
SR intervals with those obtained from all target trials in each
patient. A very similar shape was found in all of the cases.
Nine selected examples are presented in Figs. 1–3; compare
left (A) and middle (B1) columns of curves.
P3-like waveforms were identified in 180 sites in 17
patients diffusely in the frontal, temporal, and parietal lobes
Fig. 4. Three examples of P3 waveforms recorded with the scalp CPz electrodes. Patient DJ, P3 waveform with the latency of 477 ms time-locked to the
stimulus; patient FP, P3 waveform with the latency of 484 ms time-locked to the motor response; patient HZ, P3 waveform with the latency of 562 ms with
ambiguous time relationship to stimulus and motor response. (A) ERPs to target (thick line) and non-target (thin line) stimulus, unsorted averages. (B1, B2)
Target ERPs from the same sites and patients averaged from the subgroup of trials with short SR intervals (thick line) and long SR intervals (thin line), with the
stimulus onset as a trigger (on the left, B1) and the button-pressing as a trigger (on the right, B2). The peak P3 latency difference between the two subgroups
averaged with the stimulus as a trigger was 7 ms (DJ), 55 ms (FP), 47 ms (HZ), whereas averaged with the button-pressing as a trigger it was 62 ms (DJ), 0 ms
(FP) and 24 ms (HZ). Symbol : indicates P3 waveform.
Table 2
Number of target trials, mean SR intervals (ms) and standard deviations (ms) in each patient; difference between the mean SR intervals of subgroups with long
and short SR intervals (ms)
Patient ID All target trials Subgroups of target trials Difference between
mean long and
mean short SR
intervals (ms)
Short SR interval Long SR interval
n Mean SR int.GSD
(ms)
n Mean SR int.GSD
(ms)
n Mean SR int.GSD
(ms)
BS 50 429G29.7 19 413G9.6 21 448G10.9 35
DJ 46 511G55.8 17 468G11.9 17 525G15.2 57
DM 43 606G71.8 15 546G9.0 14 622G11.5 76
FP 49 523G64.2 13 489G14.0 21 532G11.9 43
HZ 30 665G114.4 10 566G14.5 9 650G14.2 84
HP 52 575G157.0 18 479G15.1 14 547G15.9 68
JI 55 602G156.8 17 512G13.4 13 560G15.0 48
KM 49 565G58.3 15 514G12.8 13 566G14.7 52
KP 48 533G84.9 14 477G16.8 14 577G17.5 100
KJ 51 470G38.9 24 467G10.1 13 519G12.1 52
LL 56 587G91.5 14 514G13.6 16 582G12.3 68
PM 60 506G55.0 18 472G8.7 17 512G10.0 40
SM 57 477G50.7 16 438G15.4 22 480G8.7 42
SO 53 446G44.0 16 407G12.4 16 457G11.1 50
SI 61 587G87.9 22 547G10.9 14 598G12.2 51
VH 55 521G67.7 13 470G13.5 21 540G13.8 70
VP 53 458G50.1 16 418G11.1 15 483G13.7 65
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171 165
structures (9 selected examples are presented in Figs. 1–3,
section A). The mean percentage of task performance errors
taken across patients was 1.2G0.7%.
The comparison of the differences of peak latencies
between the two subgroups, averaged with the stimulus
onset as a trigger and with a button-pressing as a trigger,
made it possible to distinguish 3 different types of
intracerebral P3-like waveforms.
(1) P3-like waveforms time-locked to the stimulus (3
examples in 3 patients in the parahippocampal gyrus (KJ),
the anterior cingulate gyrus (SI), and the orbitofrontal cortex
(VH) are shown in Fig. 1).
When averaged with the stimulus onset as a trigger, these
responses exhibited identical P3 peak latencies in the short
and long SR interval averages. Using the averaging with a
button-pressing as a trigger, the peak latencies differed
between the two subgroups (they were longer in the long SR
interval averages).
This type of P3-like waveforms was identified in 30 sites
in 11 patients (a summary of their intracerebral distribution
is presented in Table 3). The peaks were located on the time
axis at the point corresponding to 71.1G12.7% of the mean
SR interval (393.7G81.5 ms in absolute values).
(2) P3-like waveforms time-locked to the motor response
(3 examples in 3 patients in the posterior cingulate gyrus
(SM), the rectus gyrus (PM) and the premotor cortex (HP)
are shown in Fig. 2).
When averaged with a button-pressing as a trigger, these
responses exhibited identical P3 peak latencies in the short
and long SR interval averages. Using the averaging with the
stimulus as a trigger, the peak latencies differed between
the two subgroups (they were longer in the long SR interval
averages).
This type of P3-like waveforms was identified in 52 sites
in 13 patients (a summary of their intracerebral distribution
is presented in Table 3). The peaks were located on the time
axis at the point corresponding to 85.5G7.9% of the mean
SR interval (439.6G53.6 ms in absolute values).
(3) P3-like waveforms with ambiguous time relationship
to stimulus and motor response (3 examples in 3 patients in
the hippocampus posterior (HZ), the putamen (SO), and the
ventrolateral prefrontal cortex (VH) are shown in Fig. 3).
For the third type of P3-like waveforms, the peak P3
latency differences between the two subgroups obtained by
averaging the stimulus and button-pressing as triggers were
32.5G16.0 and 28.3G15.2 ms, respectively.
This type of P3-like waveform was identified in 98 sites
in 16 patients (a summary of their intracerebral distribution
is presented in Table 3). The peaks were located on the time
axis at the point corresponding to 78.1G12.9% of the mean
SR interval (425.3G82.2 ms in absolute values).
The one-way ANOVA revealed the main effect for the
peak P3 latencies expressed as the percentage value of
the corresponding mean SR interval, i.e. 71.1G12.7% for
the first type of P3-like waveform, 85.5G7.9% for the
second type of P3-like waveform, and 78.1G12.9% for the
third type of P3-like waveform (F (2,177)Z14.89,
P!0.000). A post hoc Spjotvoll/Stoline test showed
statistically significant differences between the first and
the second type of P3-like waveform (P!0.000) and the
second and the third types of P3-like waveform (P!0.004).
The one-way ANOVA revealed the main effect also for
Table 3
Total number of investigated sites in each structure and intracerebral distribution of 3 types of the P3-like waveform
Structures Investigated sites
hemisphere left/right
P3-like waveforms time-locked to P3-like waveforms with ambiguous
time relationship to stimulus and
motor response (sites/patients)
Stimulus
(sites/patients)
Motor response
(sites/patients)
Cingulate gyrus 10/18 1L 3R/4 2Ra
/2 1L 6R/5
Parahippocampal gyrus 18/22 1L 1R/2 0 4L 2R/5
Hippocampus 33/38 2R/2 6R/4 3L 8R/5
Amygdala 28/16 0 1L 1R/2 3L 3R/5
Basal ganglia 23/23 2R/1 2L/2 4L 2R/4
Temporal gyri (superior, middle, inferior) 64/78 3L 4R/6 3L 12R/9 12L 8R/9
Fusiform gyrus 6/10 0 3R/3 2L 1R/3
Angular gyrus 2/0 0 0 1L/1
Sensorimotor cortex 0/10 0 3R/2 2R/2
Dorsolateral prefrontal cortex 23/15 3L 1R/1 2L 5R/5 14L 7R/5
Basal and medial frontal cortex 28/26 2L 1R/2 1L 6R/2 5L 6R/6
Parietal cortex 4/16 2R/1 1L 1R/1 2R/2
White matter
Frontal lobe 9/19 2L 1R/2 1L 1R/2 3L 2R/4
Temporal lobe 5/14 1R/1 0 1L 2R/3
Parietal lobe 0/3 0 1R/1 0
L, R indicate the left and right hemispheres.
a
In the cingulate gyrus one site from two was found in left handed patient PM.
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171166
the peak P3 latencies expressed in absolute values, i.e.
393.7, 439.6, and 425.3 ms, respectively (F (2,177)Z3.54,
P!0.031). A post hoc Spjotvoll/Stoline test showed
statistically significant differences between the first and
the second type of P3-like waveforms (P!0.049).
In a number of patients we could identify more then one
P3-like waveform of each type with different latencies
diffusely distributed within the brain. For example, in
patient HP we found P3-like waveforms time-locked to the
stimulus in the inferior temporal gyrus (312 ms), in the
superior temporal gyrus (383 ms), in the parahippocampal
gyrus (305 ms) and in the cingulate gyrus (430 ms); P3-like
waveform time-locked to the motor response in the
premotor cortex (398 ms); P3-like waveform with ambiguous
time relationship to stimulus and motor response in the
cingulate gyrus (391 ms), in the frontoorbital cortex
(336 ms), in the fusiform gyrus (297 ms), in the parahippocampal
gyrus (328 ms), in the inferior temporal gyrus
(297 ms) and in the middle temporal gyrus (328 ms).
Furthermore, the localization of electrodes in each patient
was not identical. Even as we describe the position of
certain contact in several patients, for example in the
hippocampus, in each patient these contacts can record a
signal from different functional populations of neurons.
That is why the ANOVA was applied for statistical analysis
of the differences between the 3 types of P3-like waveforms
and it used all 180 sites as independent samples.
For each type of P3-like waveform, the differences of
peak P3 latencies between the long and short SR interval
averages were correlated with the differences of SR
intervals between the long and short subgroups. For the
first type of P3-like waveform, time-locked to the stimulus,
the correlation coefficient was 0.132 (P!0.485; NZ30); for
the second type of P3-like waveform, time-locked to the
motor response, the correlation coefficient was 0.784 (P!
0.000; NZ52); for the third type of P3-like waveform, with
ambiguous time relationship to stimulus and motor
response, the correlation coefficient was 0.278 (P!0.005;
NZ98).
The two most common intracerebral patterns, the
triphasic N2a–P3a-SW (see Fig. 2, column A, patient PM)
and the broad P3b (see Fig. 1, column A, patient KJ), were
identified in all the 3 types of P3-like waveforms. In the
groups of P3-like waveforms time-locked to the stimulus
and with an ambiguous time relationship to stimulus and
motor response, they were found approximately in a 3:1
ratio, whereas in the group of P3-like waveforms timelocked
to the motor response, they were present almost
equally.
Records from CPz electrodes were obtained only in 14
subjects. Due to the number of target trials in subgroups
with short and long SR intervals, which was mostly too
small for scalp averaging, we could reliably identify and
analyze the P3 waveform only in 5 subjects. In patients DJ
and JI it was time-locked to the stimulus (with the latencies
of 477 and 508 ms, respectively), in patient FP it was
time-locked to the motor response (with the latency of
484 ms) and in patients HZ and PM it represented P3
waveform with an ambiguous time relationship to stimulus
and motor response (with the latencies of 562 and 555 ms,
respectively). Examples of scalp recorded P3 waveforms are
presented in Fig. 4.
4. Discussion
4.1. Methodological remarks about intracerebral EEG
Like scalp EEG, intracranial (iEEG) has a very high
temporal resolution, limited only by sampling frequency (in
this study, 8 ms). Unlike scalp EEG, iEEG also has very
high spatial resolution, limited only by the electrode size
and spacing. However, depth electrodes are implanted only
into a limited region for strictly clinical purposes in patients
with medically intractable epilepsies and, thus, they may
record abnormal responses. Nonetheless, it is possible to
select sites and epochs for analyses that are electrographically
normal. Furthermore, the localization of electrophysiological
correlates is not really accurate until it is
supported by finding its local generator. These facts have to
be considered when interpreting intracerebrally recorded
data.
Due to the 8 ms temporal resolution, we were limited in
our separation of the P3-like waveforms into the 3 types
according to their latency difference, which could be 0, 8,
16, 24 ms and so on. In light of the measured latency
differences, we created the rule that the latency difference of
0 or 8 ms, when averaged from the stimulus onset as a
trigger or button-pressing as a trigger, means that the
waveform is time-locked to the stimulus or the motor
response, respectively. It is clear that using a higher
sampling frequency would probably provide slightly
different data separation into all the 3 types of P3-like
waveforms.
A unique aspect of this study has been the finding that a
smaller number of EEG epochs is sufficient for averaging to
produce the depth P3-like waveform. The recommended
minimum for the scalp ERPs is 20 single epochs (Cohen and
Polich, 1997; Polich, 1986a,b). In our analysis of the
relationship between the P3-like waveform and the length of
SR interval, we averaged from 9 to 24 EEG epochs and
obtained curves basically identical with those averaged
from all the recorded epochs (approximate mean number of
50). Similarly, a small number of trials was sufficient in
intracerebral error-related potentials recording in our
previous study (Bra´zdil et al., 2002). It can be explained
by the presumption that the small region surrounding the
electrode contacts provides distinctive responses, and the
signal is recorded with higher signal-to-noise ratio than is
the case with scalp-recorded ERPs. These matters have been
treated more formally by Effern et al. (2000). This fact
enabled the selection of trials with shorter and longer SR
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171 167
intervals and to create two individual groups, which could
then be averaged and compared separately in each patient.
To classify P3 potentials as stimulus- and responserelated,
it is also possible to use the comparison of the peak
amplitudes between stimulus- and response-locked
averages. If the stimulus-locked amplitude is larger than
response-locked amplitude, the potential is more stimulusrelated
than response-related, and vice versa in the other
case. This analysis has already been applied to our data, and
preliminary results were published by Damborska´ et al.
(2001).
4.2. Types of P3-like waveforms
Studies in normal subjects, intracranial recording in
epileptic patients, and lesional studies in neurological
patients show that P3 does not correspond to one cognitive
process or a unitary brain potential arising from a discrete
brain region; rather, it is related to the activation of multiple
cortical and subcortical structures which are involved
during the performance of the experimental paradigms.
According to the different time relationship of P3-like
waveform to the stimulus presentation or motor response,
we were able to divide the intracerebral P3-like waveforms
into the 3 types, which could explain the sometimes
contradictory findings in scalp-recorded ERPs. Two facts
are obvious: (1) there is a diffuse, rather than a specific,
distribution of all 3 P3 types within the brain and (2) it
remains unclear how the intracerebral P3 generators
contribute to the scalp-recorded potentials.
The first type is characterized by constant latency of the
peak in relation to the stimulus occurrence and is
independent of the length of SR interval. This potential
could be probably strictly related to the stimulus detection
and evaluation processes. This type of P3-like waveform
could possibly contribute to the scalp-recorded P3 in
paradigms where, on some presented stimuli, no response
is required, for example, in a passive oddball paradigm
(Polich, 1989) or in studies where the manipulations with
the stimuli resulted in an increase of the RT as well as of the
P3 latency: the reduction of stimulus intensity (Verleger
et al., 1991), the reduction of stimulus discriminability
(Novak et al., 1990), increasing the number of irrelevant
stimuli while randomly changing the location of the relevant
stimulus (Magliero et al., 1984), presenting flanking stimuli
to which the alternative response is assigned (Smid et al.,
1990). In the 3-stimulus variant of the oddball paradigm,
there is an additional infrequent non-target stimulus inserted
into the sequence of infrequent target and frequent standard
stimuli (Katayama and Polich, 1999). Even with the absence
of a response, this unexpected and novel stimulus generates
P3a waveform which probably reflects activities engaged
during involuntary response to novel events (Knight and
Scabini, 1998). Therefore, it is also possible that the first
type of P3-like waveform time-locked to the stimulus could
partially represent an electrophysiological correlate of
the orienting response. In support of this hypothesis, this
type of P3-like waveform occurred in the N2a–P3a-SW
pattern markedly often. However, it may also contribute to
the generation of all the scalp-recorded potentials evoked by
any kind of stimuli.
The second type of P3-like waveform is time-locked to
the moment of motor response, i.e. it is dependent on the
length of SR interval. This can be explained by the
activation of structures engaged in response-related processes,
such as timing, decision to perform the movement,
preparation for, initiation and execution of the movement.
Unfortunately the design of the oddball paradigm, with only
a button-pressing response, used in our study does not allow
distinguishing between intracerebrally recorded MRPs or
P3-like potentials influenced by response-related processes.
In our previous paper (Bra´zdil et al., 2003), the effect of the
response type on the P3-like potential was investigated with
respect to button-pressing versus mental counting conditions
from the same intracerebral sites. Task-specific P3like
potentials for the button-pressing task were observed in
the lateral premotor cortex and sporadically in the
orbitofrontal, cingulate, lateral parietal, and mesiotemporal
cortex (amygdala and fusiform gyrus), but mostly without
the character of its local origin in these sites. These findings
are in accordance with the intracerebral distribution of the
second type of P3-like waveform, which could be linked to
the family of MRPs.
The influence of response execution on P3 latency was
explored during the performance of simple RT and SR
compatibility tasks. When the stimulus evaluation demands
were minimal, response execution affected P3 latency,
however, these movement effects were modest (Doucet and
Stelmack, 1999). Another study focused on the effect of
button-pressing on P3, with the result that button-pressing
generated smaller P3 than silent-counting, and P3 topography
in button-pressing tasks was confounded by motor
potentials (Salisbury et al., 2001). As in the study by Kok
(1988), they assumed that motor processes on a simple RT
task were roughly equivalent to a ‘go’ trial in an oddball
task. The existence of the second type of P3-like waveform
demonstrated in our study could possibly explain these
findings.
The third type of P3-like waveform, which is not strictly
time-locked to stimulus or motor response, could be
interpreted as an integration of stimulus- and responseprocessing
(more precisely, as a decision to respond on the
basis of perceptual analysis) and a representation of
attentional and memory processes. The scalp-recorded P3b
with the central/parietal maximum is associated with these
processes (McCarthy et al., 1997). The P3b reductions in
patients with temporoparietal lesions are accompanied with
attention and memory deficits (Woods et al., 1993).
Temporoparietal lesions in monkeys also result in auditory
memory deficits (Colombo et al., 1990). Ruchkin et al.
(1990, 1992) provided evidence that longer latency scalp
positivities index activity in phonologic and visuospatial
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171168
systems of working memory. Our finding of this third type
of P3-like waveform in structures such as the hippocampus,
parahippocampal gyrus, amygdala and others can further
corroborate the interpretation above. It is possible that
decision making, and attentional and memory processes
need not be strictly time-locked to the stimulus occurrence,
and their time relationship can vary from trial to trial
depending on instantaneous conditions. The same reasoning
can also be used for mental counting processes required in
the task, which can contribute to generation of the third type
of P3-like waveform.
Only 5 cases of our CPz recordings yielded P3
waveform, which could be identified and analyzed in both
short and long averages. Two of them (patients DJ and JI)
would match up with the first type, one case (patient FP)
with the second type and the last two cases (patients HZ and
PM) with the third type of P3-like waveform. In patient JI,
FP and PM we found simultaneously all the 3 types of P3like
waveforms with different latencies in the various brain
structures. In patients DJ and HZ we did not find any
stimulus-related P3-like waveform. Limited number of CPz
records did not allow any detailed analysis of the correlation
between scalp-recorded and intracerebrally recorded P3
waveforms.
The presence of both kinds of typical intracerebral
patterns, the triphasic N2a–P3a-SW and the broad P3b, in
all 3 types of P3-like waveforms evokes the explanation that
these patterns could more likely represent a distinctive
electrophysiological rule, when they are recorded directly
from the nervous tissue, than a certain linkage to specific
cognitive processes.
4.3. P3-like waveform and the time axis
Our interpretation of each P3-like waveform type can be
further supported by the fact that the first type occurs on the
average on the time axis more closely to the stimulus,
the second type more closely to the motor response, and the
third type somewhere between them, i.e. 394, 440, and
425 ms, respectively. If we assume that the first type of
depth P3 can be similar to the scalp P3a, and that the third
type of depth P3 could be similar to the scalp P3b, then the
difference of 31 ms between the two types does not really
correspond to the difference of 60–80 ms between scalp P3a
and P3b (Knight and Scabini, 1998). On the other hand, it is
found that intracranial putative generators of scalp-recorded
P3 do not have the same latencies as corresponding scalp
P3a and P3b. It is also obvious that we could not register
EEG responses from all the possible intracerebral generators,
and thus our latency measurement is limited.
To determine the relation between the P3 latency and the
length of SR interval for each type of P3-like waveform, the
differences of peak P3 latencies between the long and short
SR intervals averages were correlated with the differences
of SR intervals between the long and short subgroups. No
correlation for the first type of P3-like waveform was
expected because the peak latency difference of 0–8 ms was
selected as a criterion for this type of response. For the
second type of P3-like waveform, the correlation coefficients
showed that the peak latency was dependent on the
length of SR interval, which was less evident for the third
type P3-like waveform.
4.4. Other relevant intracerebral studies
Kropotov and Ponomarev (1991) recorded the ERPs in
visual oddball paradigm from the globus pallidus and
ventro-lateral nucleus of the thalamus in parkinsonian
patients. In both structures, they identified 3 different
neuronal populations: stimulus-bound, response-bound, and
group of neurons which did not react significantly in an
oddball task. Such stimulus- and response-bound neuronal
populations were involved in generation of P3-like waveforms.
The authors included these results in their hypothesis
of action programming; P3-like activity was considered as
an electrophysiological reflection of several mental operations
(Kropotov, 1989). Turak et al. (2002) investigated the
cingulate gyrus, which was activated during various
experimental tasks. They concluded that this area appears
to be multimodal, and is involved in several types of
cognitive activity. In accordance with this conclusion, our
study showed all 3 types of P3-like waveform in this area,
too, as well as in the superior, middle and inferior temporal
gyri, hippocampus, dorsolateral prefrontal, basal and medial
frontal and parietal cortices. Our findings extend recent
knowledge that not only the cingulate gyrus, globus
pallidus, and ventro-lateral nucleus of the thalamus, but
also some areas and structures in the frontal, temporal, and
parietal lobes are involved in several type of processes
engaged in an oddball paradigm.
Our results suggest that intracerebral P3-like waveform
represents different processes diffusely distributed within
the brain. In addition to higher cognitive processes, e.g.
stimulus identification or decision-making, there must be
brain structures responsible for planning, preparation, and
execution of the response given by the task. It seems that P3like
waveform could be either stimulus-related or responserelated
or not strictly related to the stimulus or motor
response. Therefore our results support the hypothesis that
the P3 waveform registered by surface electrodes represents
a heterogeneous phenomenon.
Acknowledgements
We wish to thank J. Hala´mek for his assistance and
support. This research was supported by Research Project
MSˇMT Cˇ R 28 11 01, and by grant no. 102/02/1339 of the
Grant Agency of the Czech Republic.
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171 169
References
Baudena P, Halgren E, Heit G, Clarke JM. Intracerebral potentials to rare
target and distractor auditory and visual stimuli. III. Frontal cortex.
Electroenceph Clin Neurophysiol 1995;94:251–64.
Bennington JY, Polich J. Comparison of P300 from passive and active
tasks for auditory and visual stimuli. Int J Psychophysiol 1999;34:
171–7.
Bra´zdil M, Rektor I, Dufek M, Daniel P, Jura´k P, Kuba R. The role of
frontal and temporal lobes in visual discrimination task—depth ERP
studies. Neurophysiol Clin 1999;29:339–50.
Bra´zdil M, Rektor I, Daniel P, Dufek M, Jura´k P. Intracerebral event-related
potentials to subthreshold target stimuli. Clin Neurophysiol 2001;
112(4):650–61.
Bra´zdil M, Roman R, Falkenstein M, Daniel P, Jura´k P, Rektor I. Error
processing—evidence from intracerebral ERP recordings. Exp Brain
Res 2002;146:460–6.
Bra´zdil M, Roman R, Daniel P, Rektor I. Intracerebral somatosensory
event-related potentials: effect of response type (button pressing versus
mental counting) on P3-like potentials within the human brain. Clin
Neurophysiol 2003;114(8):1489–96.
Cohen J, Polich J. On the number of trials needed for P300. Int
J Psychophysiol 1997;25:249–55.
Colombo M, D’Amato MR, Rodman HR, Gross CG. Auditory association
cortex lesions impair auditory short-term memory in monkeys. Science
1990;247:3336–8.
Comerchero MD, Polich J. P3a, perceptual distinctiveness, and stimulus
modality. Cogn Brain Res 1998;7:41–8.
Comerchero MD, Polich J. P3a and P3b from typical auditory and visual
stimuli. Clin Neurophysiol 1999;110:24–30.
Cracco RQ, Cracco JB. Somatosensory evoked potentials in man: far-field
potentials. Electroenceph Clin Neurophysiol 1976;41:460–6.
Damborska´ A, Bra´zdil M, Jura´k P, Roman R, Kukleta M. Steep U-shaped
EEG potentials preceding the movement in oddball paradigm: their role
in movement triggering. Homeostasis 2001;41:60–3.
Donchin E, Ritter W, McCallum WC. Cognitive psychophysiology: the
endogenous components of the ERP. In: Callaway E, Tueting P,
Koslow SH, editors. Event-related brain potentials in man. New York:
Academic Press; 1978. p. 349–411.
Doucet C, Stelmack RM. The effect of response execution on P3 latency,
reaction time, and movement time. Psychophysiology 1999;36:
351–63.
Duncan-Johnson CC. P300 latency: a new metric of information
processing. Psychophysiology 1981;18:207–15.
Effern A, Lehnertz K, Fernandez G, Grunwald T, David P, Elger CE. Single
trial analysis of event related potentials: non-linear de-noising with
wavelets. Clin Neurophysiol 2000;111(12):2255–63.
Falkenstein M, Hohnsbein J, Hoorman J. Effect of choice complexity on
different subcomponents of the late positive complex of the eventrelated
potentials. Electroenceph Clin Neurophysiol 1994a;92:
148–60.
Fitzpatrick P, Klorman R, Brumaghim JT, Keefover RW. Effect of
methylphenidate on stimulus evaluation and response processes:
evidence from performance and event-related potentials. Psychophysiology
1988;25:292–304.
Halgren E, Baudena P, Clarke JM, Heit G, Lie´geois C, Chauvel P,
Musolino A. Intracerebral potentials to rare target and distractor
auditory and visual stimuli. I. Superior temporal plane and parietal lobe.
Electroenceph Clin Neurophysiol 1995a;94:191–220.
Halgren E, Baudena P, Clarke JM, Heit G, Marinkovic K, Deveaux B,
Vignal J, Biraben A. Intracerebral potentials to rare target and distractor
auditory and visual stimuli. II. Medial, lateral and posterior temporal
lobe. Electroenceph Clin Neurophysiol 1995b;94:229–50.
Halgren E, Marinkovic K, Chauvel P. Generators of the late cognitive
potentials in auditory and visual oddball tasks. Electroenceph Clin
Neurophysiol 1998;106:156–64.
Hillyard SA, Picton TW, Regan D. Sensation, perception and attention:
analysis using ERPs. In: Callaway E, Tueting P, Koslow SH, editors.
Event-related brain potentials in man. New York: Academic Press;
1978. p. 223–321.
Jentzsch I, Sommer W. Sequence-sensitive subcomponents of P300:
topographical analyses and dipole source localization. Psychophysiology
2001;38(4):607–21.
Jeon YW, Polich J. P3a from a passive visual stimulus task. Clin
Neurophysiol 2001;112:2202–8.
Jewett DL, Romano MN, Williston JS. Human auditory evoked potentials:
possible brainstem components detected on the scalp. Science 1970;
167:1517–8.
Katayama J, Polich J. Auditory and visual P300 topography from a 3
stimulus paradigm. Clin Neurophysiol 1999;110:463–8.
Knight RT, Scabini D. Anatomic bases of event-related potentials and their
relationship to novelty detection in humans. Clin Neurophysiol 1998;
15(1):3–13.
Kok A. Overlap between P300 and movement-related potentials: a response
to Verleger. Biol Psychol 1988;27:51–8.
Kropotov JD. Functional organization of memory and perception: action
programming hypothesis (In Russian). Human Physiol 1989;15:19–27.
Kropotov JD, Ponomarev VA. Subcortical neuronal correlates of component
P300 in man. Electroenceph Clin Neurophysiol 1991;78:40–9.
Kutas M, McCarthy G, Donchin E. Augmenting mental chronometry: the
P300 as a measure of stimulus evaluation. Science 1977;197:792–5.
Magliero A, Bashore TR, Coles MGH, Donchin E. On the dependence of
P300 latency on stimulus evaluation processes. Psychophysiology
1984;21:171–86.
McCarthy G, Donchin E. A metric for thought: a comparison of P300
latency and reaction time. Science 1981;211:77–80.
McCarthy G, Luby M, Gore J, Goldman-Rakic P. Infrequent events
transiently activate human prefrontal and parietal cortex as measured by
functional MRI. J Neurophysiol 1997;77:1630–4.
Mertens R, Polich J. P300 from a single-stimulus paradigm: passive versus
active tasks and stimulus modality. Electroenceph Clin Neurophysiol
1997;104:488–97.
Nandrino JL, El Massioui F. Temporal localization of the response
selection processing stage. Int J Psychophysiol 1995;19:257–61.
Novak GP, Ritter W, Vaughan HG, Wiznitzer ML. Differentiation of
negative event-related potentials in an auditory discrimination task.
Electroenceph Clin Neurophysiol 1990;75:255–75.
Paller KA, Kutas M, Mayes AR. Neural correlates of encoding in an
incidental learning paradigm. Electroenceph Clin Neurophysiol 1987;
67:360–71.
Pfefferbaum A, Christensen C, Ford JM, Kopell BS. Apparent response
incompatibility effects on P300 latency depend on the task. Electroenceph
Clin Neurophysiol 1986;64:424–37.
Polich J. Attention, probability, and task demands as determinants of P300
latency from auditory stimuli. Electroenceph Clin Neurophysiol 1986a;
63:251–9.
Polich J. P300 development from auditory stimuli. Psychophysiology
1986b;23:590–7.
Polich J. P300 from a passive auditory paradigm. Electroenceph Clin
Neurophysiol 1989;74(4):312–20.
Polich J. P300 clinical utility and control of variability. J Clin Neurophysiol
1998;15(1):14–33.
Rektor I, Louvel J, Lamarche M. Intracerebral recording of potentials
accompanying simple limb movements: a SEEG study in epileptic
patients. Electroenceph Clin Neurophysiol 1998;107:277–86.
Rektor I, Kanˇovsky´ P, Baresˇ M, Bra´zdil M, Streitova´ H, Klajblova´ H,
Kuba R, Daniel P. A SEEG study of ERP in motor and premotor
cortices and in the basal ganglia. Clin Neurophysiol 2003;114:463–71.
Ruchkin DS, Johnson Jr R, Canoune HL, Ritter W, Hammer M. Multiple
sources of P3b associated with different types of information.
Psychophysiology 1990;27:157–76.
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171170
Ruchkin DS, Johnson Jr R, Grafman J, Canoune HL, Ritter W. Distinctions
and similarities among working memory processes: an event-related
potential study. Cogn Brain Res 1992;1:53–66.
Salisbury DF, Rutherford B, Shenton ME, McCarley RW. Button-pressing
affects P300 amplitude and scalp topography. Clin Neurophysiol 2001;
112:1676–84.
Smid HGOM, Mulder G, Mulder LJM. Selective response activation can
beginbeforestimulus recognitioniscomplete: a psychophysiologicaland
error analysis of continuous flow. Acta Psychologica 1990;74:169–201.
Squires NK, Squires KC, Hillyard SA. Two varieties of long-latency
positive waves evoked by unpredictable auditory stimuli in man.
Electroenceph Clin Neurophysiol 1975;38:387–401.
Sutton S, Barren M, Zubin J, John ER. Evoked potentials correlates of
stimulus uncertainty. Science 1965;150:1187–8.
Turak B, Louvel J, Buser P, Lamarche M. Event-related potentials
recorded from cingulate gyrus during attentional tasks: a study in
patients with implanted electrodes. Neuropsychologia 2002;40:
99–107.
Vaughan Jr HG, Costa LD, Ritter W. Topography of the human motor
potential. Electroenceph Clin Neurophysiol 1968;25:1–10.
Verleger R. On the utility of P3 latency as an index of mental chronometry.
Psychophysiology 1997;34:131–56.
Verleger R, Neuka¨ter W, Ko¨mpf D, Vieregge P. On the reasons for the
delay of P3 latency in healthy elderly subjects. Electroenceph Clin
Neurophysiol 1991;79:488–502.
Woods Dl, Knight RT, Scabini D. Anatomical substrates of auditory
selective attention: behavioral and electrophysiological effects of
temporal and parietal lesions. Cogn. Brain Res. 1993;1:227–40.
R. Roman et al. / Clinical Neurophysiology 116 (2005) 160–171 171
Příloha č. 2
Roman R, Brázdil M, Chládek J, Rektor I, Jurák P, Světlák M, Damborská A, Shaw DJ,
Kukleta M. Hippocampal negative event-related potential recorded in humans during a simple
sensorimotor task occurs independently of motor execution. Hippocampus 2013;23(12):1337-
1344. IF (2013) = 4,302
Tato práce získala v roce 2014 ocenění České společnosti pro klinickou neurofyziologii ČLS
JEP - 1.místo v soutěži o nejlepší vědeckou práci za rok 2013.
Hippocampal Negative Event-Related Potential Recorded in Humans
During a Simple Sensorimotor Task Occurs Independently of Motor
Execution
Robert Roman,1,2
* Milan Brazdil,2,3
Jan Chladek,2,4
Ivan Rektor,2,3
Pavel Jurak,4
Miroslav Svetlak,1
Alena Damborska,1,2
Daniel J. Shaw,2
and Miloslav Kukleta2
ABSTRACT: A hippocampal-prominent event-related potential (ERP)
with a peak latency at around 450 ms is consistently observed as a correlate
of hippocampal activity during various cognitive tasks. Some intracranial
EEG studies demonstrated that the amplitude of this hippocampal
potential was greater in response to stimuli requiring an overt motor
response, in comparison with stimuli for which no motor response is
required. These findings could indicate that hippocampal-evoked activity
is related to movement execution as well as stimulus evaluation and associated
memory processes. The aim of the present study was to investigate
the temporal relationship between the hippocampal negative potential
latency and motor responses. We analyzed ERPs recorded with 22 depth
electrodes implanted into the hippocampi of 11 epileptic patients. Subjects
were instructed to press a button after the presentation of a tone.
All investigated hippocampi generated a prominent negative ERP peaking
at ~420 ms. In 16 from 22 cases, we found that the ERP latency did not
correlate with the reaction time; in different subjects, this potential could
either precede or follow the motor response. Our results indicate that the
hippocampal negative ERP occurs independently of motor execution. We
suggest that hippocampal-evoked activity, recorded in a simple sensorimotor
task, is related to the evaluation of stimulus meaning within the context
of situation. VC 2013 Wiley Periodicals, Inc.
KEY WORDS: intracranial recordings; auditory task; hippocampus;
ERP latency; motor response
INTRODUCTION
Electrophysiological recordings from the hippocampus in humans
(e.g., Halgren et al., 1980; Stapleton and Halgren, 1987; Grunwald
et al., 1995; Brazdil et al., 2001; Fell et al., 2005; Boutros et al., 2008;
Axmacher et al., 2010) and in animals (e.g., Paller
et al., 1992; Shinba et al., 1996; Shin, 2011) reveal a
large, mostly negative, event-related potential (ERP)
with a peak latency of around 450 ms. There is no
consensus on the terminology used to refer to this
cognitive potential. Some authors regard it as one of
the putative generators of the scalp recorded P300
and label this ERP accordingly; Grunwald et al.
(1999) and Fell et al. (2004), for example, used the
term MTL-P300, while Halgren et al. (1995) labeled
it the depth P3b. Likewise, similar ERPs recorded
from the hippocampus and various other brain structures
have been referred to as P3-like waveforms
(Brazdil et al., 2003; Roman et al., 2005; Damborska
et al., 2012). Hippocampal activity is not related to
some functions associated with scalp P300, however,
and it can be observed also after frequent stimuli in
oddball task (Kukleta et al., 2003). We believe that
the descriptive term “large negative ERP” used by Paller
and McCarthy (2002) seems to be the most accurate
so far, and hereafter, we refer to this potential as
the hippocampal negative ERP, or simply the negative
potential, rather than the MTL-P300 or the P3-like
waveform.
McCarthy et al. (1989) have speculated that this
hippocampal ERP is generated by the synchronous
activation of spatially aligned hippocampal pyramidal
cells. Later, these same authors suggested that this
potential may reflect the arrival into the hippocampus
of neocortical information, of modulatory inputs
from diffusely projecting brain systems, or of recurrent
inhibition from hippocampal pyramidal cells
(Paller and McCarthy, 2002). As such, this ERP has
been used as an index of hippocampal activity during
various cognitive tasks.
One view on the medial temporal lobe functional
arrangement is the hierarchically organized set of associational
networks. Associational connections within
the perirhinal, parahippocampal, and entorhinal cortices
enable a significant amount of integration of
unimodal and polymodal inputs, so that only highly
integrated information reaches the remainder of the
hippocampal formation (Lavenex and Amaral, 2000).
Sensory inputs to the long axis of the hippocampus
are organized topographically, as are its efferent connections.
Studies report an anterior–posterior
1
Department of Physiology, Medical Faculty, Masaryk University, Brno,
Czech Republic; 2
CEITEC, Central European Institute of Technology,
Masaryk University, Brno, Czech Republic; 3
Department of Neurology,
St. Anne Hospital, Masaryk University, Brno, Czech Republic; 4
Institute
of Scientific Instruments, Academy of Sciences of the Czech Republic,
Brno, Czech Republic
Grant sponsor: European Regional Development Fund and MSMT CR;
Grant numbers: CZ.1.05/1.1.00/02.0068, MSM0021622404; CZ.1.05/
2.1.00/01.0017; Grant sponsor: GACR Czech Science Foundation; Grant
number: GAP103/11/0933.
*Correspondence to: Robert Roman, Department of Physiology, Masaryk
University, Kamenice 5, Brno 625 00, Czech Republic. E-mail:
roman@med.muni.cz
Accepted for publication 17 July 2013.
DOI 10.1002/hipo.22173
Published online 9 September 2013 in Wiley Online Library
(wileyonlinelibrary.com).
VC 2013 WILEY PERIODICALS, INC.
HIPPOCAMPUS 23:1337–1344 (2013)
differentiation of hippocampal activation during various tasks
(Paller and McCarthy, 2002; Crottaz-Herbette et al., 2005;
Ludowig et al., 2010). Such functional organization is likely
based on interconnections between transverse circuits within
the longitudinal axis of hippocampus (Small, 2002; Aggleton,
2010; Aggleton et al., 2012).
The hippocampus is involved in novelty detection (Grunwald
et al., 1998; Cohen et al., 1999; Strange and Dolan,
2001), salience detection (Rosburg et al., 2007), spatial memory
and navigation (e.g., Watrous et al., 2011), and in encoding
and recollection of episodic memory (e.g., Fernandez el al.,
2002; Eichenbaum, 2004; Amaral and Lavenex, 2007; Ranganath,
2010; Lega et al., 2012; Eichenbaum et al., 2012). In a
recently published review, Olsen et al. (2012) suggest that the
hippocampus supports multiple cognitive processes through
relational binding and comparison, with or without conscious
awareness for the relational representations that are formed,
retrieved, and/or compared.
Despite the fact that most evidence points to cognition variables,
some animal studies showed hippocampal theta activity
associated with motor behavior (Vanderwolf, 1969; Wyble
et al., 2004; Bland et al., 2006; Shin, 2011). Intracranial
recordings from human hippocampi have also demonstrated
movement-related increase in theta oscillatory activity (Ekstrom
et al., 2005). These findings are consistent with sensorimotor
integration theory, according to which hippocampal theta activity
may reflect both the processing of sensory information
related to the preparation for movement and the execution of
movement (Bland and Oddie, 2001).
Besides theta oscillatory power changes, additional evidence
for the role of the hippocampus in motor execution comes
from studies demonstrating a modulation of hippocampalevoked
activity during motor responses. In studies employing
the oddball task, for example, the amplitude of the hippocampal
ERP is consistently reported to be higher after target stimuli
requiring a motor response, relative to frequent stimuli for
which no motor response is required (e.g., Brazdil et al., 1999;
Fell et al., 2005; Ludowig et al., 2010). The amplitude of this
potential was also found to be significantly greater after button
pressing compared to mental counting (Brazdil et al., 2003).
Furthermore, in our previous study (Roman et al., 2005) we
observed that hippocampal P3-like waveforms to target stimuli
could be time-locked to the onset of the stimulus or the motor
response. These findings could indicate that this potential partially
reflects processes also related to the motor response and
not just to stimulus evaluation and associated memory functions.
To explore this possibility, we investigated the temporal
relationship between the hippocampal negative potential
latency and the onset of a stimulus-cued movement. To do so,
we employed a simple stimulus-response task, during which
subjects were required to press a single button in response to a
single, unchanging auditory stimulus. The subjects were
instructed to respond at their own pace—that is, not to
respond as fast as possible. The latency parameter of the examined
hippocampal ERPs was then analyzed in subgroups of
EEG trials with slower and faster responses.
MATERIAL AND METHODS
Subjects
Eleven patients (nine females and two males; one male subject
was left-handed; mean age 35 6 11 yrs) with medically
intractable epilepsy participated in the study. All subjects had
normal or corrected-to-normal vision. Informed consent was
obtained from each subject before the experiment, and the
study received the approval of the Ethical Committee of Masaryk
University.
Intracranial Recordings
For clinical purposes a total of 91 depth multicontact electrodes
were implanted orthogonally to the frontal, parietal, and
temporal lobes to localize seizure origin before surgical treatment.
In this study, we investigated ERP data from 22 electrodes
that were positioned at the anterior (electrodes B mostly in
hippocampal head, occasionally in the anterior part of hippocampal
body) or posterior part of the hippocampi (electrodes
C mostly in hippocampal body or its posterior part) in both
right (nine electrodes) and left hemispheres (13 electrodes).
The number of contacts located in hippocampal tissue ranged
from two to four per electrode. The number of electrodes
implanted into the hippocampi per patient varied from one to
four (see Table 1); no electrodes recorded from epileptic foci.
ERPs were recorded with standard semiflexible multilead
electrodes (ALCIS), each with a diameter of 0.8 mm. Recording
contacts were 2.0 mm in length, and successive contacts
were separated by 1.5 mm. The exact position of the electrodes
and their contacts in the brain were verified using postplacement
MRI with electrodes in situ.
The EEG was recorded with a sampling rate of 1,024 Hz
during seizure-free periods using the 128-channel TrueScan
EEG system (Deymed Diagnostic). All recordings were monopolar,
with a linked earlobe reference. All impedances were less
than 5 kX. Eye movements were recorded from a cathode
placed 1 cm lateral and 1 cm above the canthus of the left eye,
and from an anode 1 cm lateral and 1 cm below the canthus
of the right eye.
Behavioral Task
A single-stimulus auditory task was performed, consisting of
five sessions of 30 trials (i.e., 150 trials). On a given trial, we
presented a single 1-kHz tone. The duration of auditory stimuli
was constant at 200 ms. The interstimulus interval varied
randomly between 4,000 and 6,000 ms. The intersession interval
was 1 min. Subjects were instructed to press a button with
their dominant hand after presentation of the tone. They were
also instructed not to respond as fast as possible.
Data Analysis
ScopeWin and Physioplore software were used for offline analysis.
EEG data were digitally filtered with pass band from 0.1 to
1338 ROMAN ET AL.
Hippocampus
5.5 Hz. This filter was chosen because the most informative
aspect of the signal we were analyzing resides within frequencies
below 5 Hz (Jacobs et al., 2007; Lega et al., 2012; Watrous
et al., 2011, 2013). Furthermore, we wanted to exclude possible
influence of higher frequency components on the shape of the
ERPs obtained from subgroups with a low number of trials. The
filtered signal was then segmented according to the stimulation
trigger onset, and segment lengths were 2,000 ms with a 600-ms
prestimulus period used as baseline (2700 to 2100 ms before
stimulus onset). In Figures 2–4, only a short prestimulus period
(500 ms) is depicted, resulting in a shorter time axis. Segments
containing artefacts were rejected manually on the basis of visual
inspection. In each subject, the reaction time (RT) was measured
in artifact-free EEG segments. Outliers, i.e., 5% of segments
with extremely long or short RT, were excluded from subsequent
averaging. For further analysis, we selected in each electrode the
one contact with the highest peak amplitude of negative ERP.
We evaluated ERP parameters derived from fractional area
measures (Luck, 2005). We used a measurement window
delimited by a line intersecting the waveform at the 50% peak
amplitude level. The time point at which the waveform
reached 50% of its peak amplitude represented the “onset
latency.” The 50% “area latency” was obtained by computing
the area under the ERP waveform above the aforementioned
measurement window and then identifying the time point that
bisected that area. We used this latter parameter as a measure
of the latency of a given potential, and it is this particular measure
that is the focus of our analyses; for the sake of brevity,
we refer to this parameter as “ERP latency.” We also expressed
ERP latency as a relative value—i.e., the percentage fraction of
RT. This gave us a measure of “relative latency.” We measured
also the “peak latency,” allowing the comparison of our results
with the majority of other studies. Figure 1 illustrates all
parameters measured in the present study.
To investigate whether or not any temporal relationship
existed between the hippocampal negative potential and the
motor response (i.e., the moment of button pressing) we
performed the following procedure. In each subject, the
artifact-free EEG segments were sorted according to the
respective RT from the fastest to the slowest responses. Subsequently,
the segments were divided into five subgroups
and averaged separately (sorted averages). In creating the
subgroups, we attempted to fulfill the following criteria: a
sufficient number of segments for averaging in each subgroup;
an identical number of segments in all subgroups;
and similar variability of the RT across subgroups. Finally,
the ERP latency and relative latency, obtained from all five
sorted averages, were correlated with the median RT for
each subgroup.
Nonparametric statistics—Spearman correlation and Wilcoxon
pair test—were performed using the routines included
in the Statistica program (StatSoft).
TABLE 1.
Reaction Time and Hippocampal Negative ERP Parameters Measured in All Investigated Contacts, for All Subjects
Subject
Reaction time
mean 6 SD (ms)
Reaction time
median (ms) Contact
Onset
latency (ms)
50%
area latency (ms)
Peak
latency (ms)
Peak
amplitude (mV)
1 387 6 58 371 B4 300 468 512 2125
C3 324 552 528 294
B0
1 214 423 462 237
C0
2 300 503 522 292
2 660 6 61 654 B0
1 270 389 385 2136
3 260 6 31 259 B3 296 453 427 2205
C0
3 321 478 465 2239
4 384 6 64 387 C0
2 268 426 445 280
5 608 6 144 566 B2 227 369 330 2133
B0
2 226 358 339 277
C0
2 283 430 437 261
6 649 6 134 636 B0
2 239 420 361 240
7 668 6 160 652 B4 240 355 341 294
C4 226 388 407 2145
8 440 6 64 444 B3 272 430 386 2188
C3 274 442 384 2202
B0
3 295 464 494 259
9 1,113 6 269 1,117 B0
1 265 366 351 265
C0
2 279 401 383 251
10 693 6 93 688 C0
3 260 405 353 278
11 552 6 45 552 B3 267 372 370 2300
B0
4 290 434 396 2338
B 5 anterior hippocampus, C 5 posterior hippocampus, 0
5 left hemisphere.
HIPPOCAMPAL EVENT-RELATED POTENTIAL IN SENSORIMOTOR TASK 1339
Hippocampus
RESULTS
All the subjects responded to all presented stimuli. RT in a
group of 11 subjects measured only from artifact-free EEG segments
was 583 6 227 ms (mean 6 SD) and individual RT
values are listed in Table 1. The average number of artifact-free
segments across all subjects was 92 6 21, ranging from 61 to
126. The shortest responses were measured in subject 3 (260 6
31 ms), while subject 9 responded very slowly with the highest
variability of responses (1,113 6 269 ms). For the rest of the
subjects, mean RTs and standard deviations ranged, respectively,
384–693 and 45–160 ms.
We found prominent hippocampal negative potentials characterized
by a local origin, indicated by a steep voltage gradient
change, in all investigated electrodes. Two examples of locally
generated ERPs are demonstrated in Figure 2. The values of
our selected ERP parameter measurements (mean 6 SD) across
22 cases were as follows: 50% area latency 5 424 6 49 ms;
peak latency 5 413 6 62 ms; onset latency 5 270 6 31 ms;
peak amplitude 5 2129 6 84 mV. Parameter values from
each individual subject are presented in Table 1.
In all investigated contacts, averages created from sorted subgroups
of segments revealed waveforms comparable to the
waveform averaged from all artifact-free segments (compare
one example in Figs. 3A,B). The number of segments included
in each subgroup varied from 14 6 1 to 25 6 4.
To test the temporal relationship between the negative
potential latency and motor responses, in each subject, we performed
correlation analyses between two measures of ERP
latency—both absolute (50% area latency) and relative
latency—with median RT obtained from sorted subgroup averages
(Table 2). First, we observed nonsignificant correlations
between ERP latency and median RT in all but two cases; specifically,
C0
2 in subject 1 and C0
3 in subject 3. These findings
reflect relatively constant hippocampal negative potential
latency that is independent of RT. Second, we found significant
negative correlations between ERP relative latency and median
RT in 16 of 22 cases. These results demonstrate that with longer
RT the hippocampal negative potential latency, expressed as
a relative value of corresponding RT, became correspondingly
lower. Although many of the correlations observed are far more
robust than those reported typically in the literature, yet they
failed to reach statistical significance. This is unsurprising given
the low degrees of freedom. Nevertheless, we performed
resampled procedures and we obtained very similar results: for
the coefficient of 20.8 (i.e., subject 1, C0
2), P > 0.067; and
for the coefficient 20.7 (i.e., subject 1, B0
1), P 5 0.119.
In the light of the previous results, we also correlated
median RT and ERP latency parameters but measured from
FIGURE 2. Two examples of locally generated hippocampal negative ERP revealed by steep
voltage gradient changes. In subject 2 (left), contact B0
1 was located in the left hippocampal
head, while contacts B0
2–4 were located outside this structure; In subject 8 (right), contacts
C1–3 penetrate right hippocampal body, while contact C4 lay outside this structure.
FIGURE 1. Schematic representation of event-related potential
(ERP) parameters derived from fractional area measurements. The
50% area latency was used as the primary measure of ERP latency.
The time point at which the waveform reached 50% of the peak
amplitude represented the onset latency.
1340 ROMAN ET AL.
Hippocampus
unsorted averages across subjects. For this analysis, if the studied
ERP is related to the motor response one would expect
high positive correlations. On the contrary, however, we
observed significant negative correlations between median RT
and ERP latency (r 5 20.77, P < 0.001) and between
median RT and onset latency (r 5 20.55, P < 0.01). This
finding supports the proposal that the studied ERP is unrelated
to the motor response.
In all investigated sites, the hippocampal ERP latency—our
primary measure—varied along the time axis in relation to the
motor response (see Table 1). In 14 cases, it was shorter than
the median RT and preceded the motor response—in two cases
in subject 9, it was shorter by 751 and 716 ms; in 12 cases, it
varied from 297 to 2 ms (mean value 182 6 98 ms) before
the average motor response. In eight cases, the hippocampal
ERP latency was longer than the median RT; this is true especially
for subject 3, who provided the fastest responses,
whereby ERP latency was longer by 219 ms in C0
3 and 194
ms in B3. In the six other remaining cases, it varied from 20
to 181 ms (mean value 87 6 62 ms) after the motor response
(see examples in Fig. 4). In contrast, our other measure of ERP
latency—onset latency—preceded the motor response in all
interrogated contacts except two in subject 3. It also demonstrates
that there is no fixed relationship between the studied
ERP and the motor response.
Differences in the measured parameters between anterior
and posterior parts of the hippocampus were assessed in six
pairs of contacts for subjects 1, 5, 7, 8, and 9 (Wilcoxon pair
test). A significant anterior—posterior difference (P < 0.05)
revealed that ERP latency was shorter in the anterior compared
with posterior hippocampi—i.e., 400 6 47 and 452 6
63 ms, respectively. No significant differences were found
between right and left hippocampi in any of the measured
parameters in five investigated pairs of contacts (subjects 1, 5,
8, and 11).
DISCUSSION
In the present study, we found that the human hippocampi
consistently generate a large negative ERP during simple auditory
stimulus-response task. Analyses focusing on the latency of this
hippocampal ERP suggested that it is not time-locked to the
motor response (i.e., to movement execution). First, the
FIGURE 3. Hippocampal negative ERP recorded from the
right hippocampal head in subject 11, contact B3. A: ERP averaged
from 88 artifact-free EEG segments (unsorted average). B:
ERPs recorded from the same contact and patient, averaged from
five subgroups of EEG segments defined according to median RT
(sorted averages), ordered from the shortest (upper) to the longest
(lower). ERP latency is indicated by short vertical solid lines crossing
the waveforms. The vertical dashed lines depict median RT
expressed numerically on the left. C: Transversal and coronal MR
images demonstrate the location of electrode B, with the B3 contact
position indicated by a white arrow.
HIPPOCAMPAL EVENT-RELATED POTENTIAL IN SENSORIMOTOR TASK 1341
Hippocampus
hippocampal negative ERP latency measured in five sorted subgroups
of segments did not correlate with median RT. When the
same latency parameter was expressed as a relative value of RT,
however, it correlated highly with median RT in most of the
cases. Second, there is no fixed temporal relationship between the
hippocampal negative ERP and the motor response expressed as
median RT. In some cases, the ERP latency was shorter and, in
other cases, it was longer than median RT, i.e., negative potential
preceded or followed the motor response, respectively. Third,
across the entire group of subjects, the variability of the ERP
latency and the onset latency was at least four times lower than
the variability of RTs.
Our conclusion is based on a simplified conceptualization of
sensorimotor tasks; specifically, we assume that the hippocampal
ERP reflects activity related either to the stimulus presentation
or to the execution of the response. The results of our
correlation analyses between RT and the latency (absolute and
relative) of the hippocampal negative potential should then
reveal the temporal relationship between this ERP and the
stimulus or the response. In the first case, the latency of the
ERP should be independent of RT. The opposite would be
expected if the studied ERP reflects processes linked to the execution
of the response; the ERP latency should depend entirely
on RT. By averaging subgroups of EEG segments sorted
according to RT in each subject, we measured an almost constant
potential latency in all sorted averages. Moreover, when
this potential latency was expressed as a relative value of RT, it
was lower for longer RTs. In other words, we revealed that the
hippocampal ERP is linked to the auditory stimulus rather
than the motor response.
One of the suggested hippocampal functions is its support
in multiple cognitive processes through relational binding and
comparison. The comparison occurs when recently processed
perceptual information is evaluated with respect to associated
relevant information that is maintained in the memory (Vinogradova,
2001). In our study, the auditory stimulus was given a
task-relevant meaning by verbal instruction. After each presentation
of the stimulus, its identification and comparison with
memorized representation had to be accomplished. We believe
that the hippocampal negative potential could represent an
electrophysiological correlate of such evaluation processing.
In our study, several subjects performed the task so quickly
that the latency of hippocampal waveform followed the execution
of instructed movement. This could be explained by the
existence of a short-cut pathway from auditory areas to secondary
motor areas involved in movement programming (Bender
et al., 2006). This is in line with the finding of early processing
of auditory stimuli in the frontal cortex as well (Kukleta et al.,
2010). It demonstrates that aforementioned hippocampal evaluation
is not critical for movement execution.
FIGURE 4. Examples of hippocampal negative ERPs from each subject. Contacts B are
located in hippocampal head or in the anterior part of the hippocampal body, and contacts C
are positioned within the posterior part of hippocampal body; apostrophes indicate lefthemisphere
locations. ERP latency is indicated by vertical solid lines crossing the waveforms.
Shadow rectangles below each waveform represent reaction times quartile ranges and vertical
dashed lines, crossing the rectangles, signify median RT. Scaling was adjusted separately for
each waveform so as to optimize amplitude.
1342 ROMAN ET AL.
Hippocampus
Low variability of RTs in some subjects represented a certain
limitation for our analysis. In such cases, we obtained very
small differences of median RT between the subgroups we created.
Since the ERP latency varied minimally in the sorted
averages also, this may have influenced the correlation coefficients
we obtained in subjects 1 and 3.
Another finding from the present study is the primarily
shorter ERP latencies recorded from anterior relative to posterior
orthogonal electrodes; i.e., 400 6 47 and 452 6 63 ms,
respectively. The small number of cases did not allow deeper
statistical evaluation, however. Very similar but nonsignificant
anterior–posterior differences in peak latencies were also
observed in axial electrodes penetrating the hippocampal head
429 6 96 ms and hippocampal body 486 6 69 ms (Ludowig
et al., 2010). These findings seem to be in accordance with a
model of memory recall that assumes a spreading of hippocampal
activation from anterior to posterior during information
retrieval (Small, 2002). The recent finding of traveling theta
waves with the same direction of propagation observed in hippocampi
of freely behaving rats is not in contradiction with
this model (Lubenov and Siapas, 2009).
The hippocampal negative ERP was found to be uniform in
shape and polarity across all subjects and investigated sites. It is
for this reason that we chose to employ parameters derived
from fractional area measures for the description of this potential.
The 50% area latency parameter is obtained by computing
the area under the ERP waveform over a delimited measurement
window, and then identifying the time point that bisects
that area. Thus, it can be related to RT more directly because
this is similar to the median RT, which is the point separating
the fastest and slowest halves of RTs. It also has several other
advantages over peak latency that is used in the majority of
ERP studies; the same value is expected irrespective of the
noise level of the data, for example, rendering it less sensitive
to noise (Luck, 2005). That is why we use the 50% area
latency parameter as a measure of the latency of the investigated
negative potential instead of the peak latency. Concerning
RT, we use the mean RT for description of group
behavioral performance and median RT to compare the size of
the latency effect to the size of the RT effect.
The primary finding of the present study is that during a
simple sensorimotor task, human hippocampi generate a prominent
negative ERP that occurs independently of motor execution.
We suggest that this electrophysiological phenomenon is
related to evaluation of stimulus meaning within the context of
the current situation.
Acknowledgments
The authors thank Josef Halamek for his helpful comments
during the preparation of this manuscript.
REFERENCES
Aggleton JP. 2012. Multiple anatomical systems embedded within the
primate medial temporal lobe: Implications for hippocampal function.
Neurosci Biobehav Rev 36:1579–1596.
Aggleton JP, Wright NF, Vann SD, Saunders RC. 2012. Medial temporal
lobe projections to the retrosplenial cortex of the macaque
monkey. Hippocampus 22:1883–1900.
Amaral DG, Lavenex P. 2007. Hippocampal neuroanatomy. In: Andersen
P, Morris R, Amaral DG, Bliss T, O’Keefe J, editors. The Hippocampus
Book. Oxford: Oxford University Press. pp 37–114.
Axmacher N, Cohen MX, Fell J, Haupt S, D€umpelmann M, Elger
CE, Schlaepfer TE, Lenartz D, Sturm V, Ranganath CH. 2010.
Intracranial EEG correlates of expectancy and memory formation
in the human hippocampus and nucleus accumbens. Neuron 65:
541–549.
Bender S, Oelkers-AX R, Resch F, Weisbrod M. 2006. Frontal lobe
involvement in the processing of meaningful auditory stimuli
develops during childhood and adolescence. Neuroimage 33:
759–773.
Bland BH, Oddie SD. 2001. Theta band oscillation and synchrony in
the hippocampal formation and associated structures: The case for
its role in sensorimotor integration. Behav Brain Res 127:119–136.
Bland BH, Jackson J, Derrie-Gillespie D, Azad T, Rickhi A, Abriam J.
2006. Amplitude, frequency, and phase analysis of hippocampal
theta during sensorimotor processing in a jump avoidance task.
Hippocampus 16:673–681.
Boutros NN, Mears R, Pflieger ME, Moxon KA, Ludowig E, Rosburg T.
2008. Sensory gating in the human hippocampal and rhinal regions:
Regional differences. Hippocampus 18:310–316.
TABLE 2.
Spearman Correlation Coefficients (df 5 4) for the Relationship
Between Median RT and ERP Latency, and Between Median RT and
ERP Relative Latency (the Percentage Fraction of Reaction Time
Subject Contact Latency Relative latency
1 B4 0.7 20.9*
C3 0.5 20.8
B0
1 0.1 20.7
C0
2 0.9* 20.8
2 B0
1 0.7 20.1
3 B3 0.2 21.0*
C0
3 0.9* 21.0*
4 C0
2 0.6 20.9*
5 B2 0.5 20.9*
B0
2 0.5 20.9*
C0
2 20.2 21.0*
6 B0
2 0.6 20.9*
7 B4 20.3 21.0*
C4 20.3 20.9*
8 B3 20.4 20.9*
C3 20.3 20.9*
B0
3 0.1 20.6
9 B0
1 0.3 20.9*
C0
2 0.3 20.9*
10 C0
3 20.6 21.0*
11 B3 20.3 20.9*
B0
4 0.5 20.7
B 5 anterior hippocampus, C 5 posterior hippocampus, 0
5 left hemisphere.
*P < 0.05.
HIPPOCAMPAL EVENT-RELATED POTENTIAL IN SENSORIMOTOR TASK 1343
Hippocampus
Brazdil M, Roman R, Daniel P, Rektor I. 2003. Intracerebral somatosensory
event-related potentials: Effect of response type (button
pressing versus mental counting) on P3-like potentials within the
human brain. Clin Neurophysiol 114:1489–1496.
Brazdil M, Roman R, Daniel P, Rektor I. 2005. Intracerebral errorrelated
negativity in a simple Go/NoGo task. J Psychophysiol 19:
244–255.
Brazdil M, Rektor I, Dufek M, Daniel P, Jurak P, Kuba R. 1999. The
Role of Frontal and Temporal Lobes in Visual Discrimination
Task—Depth ERP Studies. Neurophysiol Clin 29: 339–350.
Brazdil M, Rektor I, Daniel P, Dufek M, Jurak P. 2001. Intracerebral
event-related potentials to subthreshold target stimuli. Clin Neurophysiol
112:650–661.
Cohen NJ, Ryan J, Hunt C, Romine L, Wszalek T, Nash C. 1999.
Hippocampal system and declarative (relational) memory: Summarizing
the data from functional neuroimaging studies. Hippocampus
9:83–98.
Crottaz-Herbette S, Lau KM, Glover GH, Menon V. 2005. Hippocampal
involvement in detection of deviant auditory and visual
stimuli. Hippocampus 15:132–139.
Damborska A, Brazdil M, Rektor I, Janousova E, Chladek J, Kukleta
M. 2012. Late divergence of target and nontarget ERPs in a visual
oddball task. Physiol Res 61:307–318.
Eichenbaum H. 2004. Hippocampus: Cognitive processes and neural representations
that underlie declarative memory. Neuron 44:109–120.
Eichenbaum H, Sauvage M, Fortin N, Komorowski R, Lipton P.
2012. Towards a functional organization of episodic memory in
the medial temporal lobe. Neurosci Biobehav Rev 36:1597–1608.
Ekstrom AD, Caplan JB, Ho E, Shattuck K, Fried I, Kahana MJ.
2005. Human hippocampal theta activity during virtual navigation.
Hippocampus 15:881–889.
Fell J, Dietl T, Grunwald T, Kurthen M, Klaver P, Trautner P Schaller
C, Elger CE, Fernandez G. 2004. Neural bases of cognitive ERPs:
More than phase reset. J Cogn Neurosci 16:1595–1604.
Fell J, K€ohling R, Grunwald T, Klaver P, Dietl T, Schaller C, Becker A,
Elger CE, Fernandez G. 2005. Phase-locking characteristics of limbic
P3 responses in hippocampal sclerosis. Neuroimage 24:980–989.
Fernandez G, Klaver P, Fell J, Grunwald T, Elger CE. 2002. Human
declarative memory formation: Segregating rhinal and hippocampal
contributions. Hippocampus 12:514–519.
Grunwald T, Elger CE, Lehnertz K, Van Roost D, Heinze HJ. 1995.
Alterations of intrahippocampal cognitive potentials in temporal
lobe epilepsy. Electroencephalogr Clin Neurophysiol 95:53–62.
Grunwald T, Lehnertz K, Heinze HJ, Helmstaedter C, Elger CE.
1998. Verbal novelty detection within the human hippocampus
proper. Proc Natl Acad Sci USA 95:3193–3197.
Grunwald T, Beck H, Lehnertz K, Bl€umcke I, Pezer N, Kutas M,
Kurthen M, Karakas HM, Van Roost D, Wiestler OD, Elger CE.
1999. Limbic P300s in temporal lobe epilepsy with and without
ammon’s horn sclerosis. Eur J Neurosci 11:1899–1906.
Halgren E, Squires NK, Wilson CL, Rohrbaugh JW, Babb TL,
Crandall PH. 1980. Endogenous potentials generated in the
human hippocampal formation and amygdala by infrequent events.
Science 210:803–805.
Halgren E, Baudena P, Clarke JM, Heit G, Marinkovic K, Devaux B,
Vignal JP, Biraben A. 1995. Intracerebral potentials to rare target and
distracter auditory and visual stimuli. II. medial, lateral and posterior
temporal lobe. Electroencephalogr Clin Neurophysiol 94:229–250.
Jacobs J, Kahana MJ, Ekstrom AD, Fried I. 2007. Brain oscillations
control timing of single-neuron activity in humans. J Neurosci 27:
3839–3844.
Kukleta M, Brazdil M, Roman R, Jurak P. 2003. Identical eventrelated
potentials to target and frequent stimuli of visual oddball
task recorded by intracerebral electrodes. Clin Neurophysiol 114:
1292–1297.
Kukleta M, Turak B, Louvel J. 2010. Very early EEG responses to a
meaningful auditory stimulus in the frontal lobes: An intracerebral
study in humans. Physiol Res 59:1019–1027.
Lavenex P, Amaral DG. 2000. Hippocampal-neocortical interaction: A
hierarchy of associativity. Hippocampus 10:420–430.
Lega BC, Jacobs J, Kahana M. 2012. Human hippocampal theta oscillations
and the formation of episodic memories. Hippocampus 22:
748–761.
Lubenov EV, Siapas AG. 2009. Hippocampal theta oscillations are
travelling waves. Nature 459:534–539.
Luck SJ. 2005. Measuring ERP latencies. In: Luck SJ, editor. An
Introduction to the Event-Related Potential Technique. Cambridge:
MIT Press. pp 237–249.
Ludowig E, Bien CG, Elger CE, Rosburg T. 2010. Two P300 generators
in the hippocampal formation. Hippocampus 20:186–195.
McCarthy G, Wood CC, Williamson PD, Spencer DD. 1989. Taskdependent
field potentials in human hippocampal formation.
J Neurosci 9:4253–4268.
Olsen RK, Moses SN, Riggs L, Ryan JD. 2012. The hippocampus
supports multiple cognitive processes through relational binding
and comparison. Front Hum Neurosci 6:1–13.
Paller KA, McCarthy G. 2002. Field potentials in the human hippocampus
during the encoding and recognition of visual stimuli.
Hippocampus 12:415–420.
Paller KA, McCarthy G, Roessler E, Allison T, Wood CC. 1992.
Potentials evoked in human and monkey medial temporal lobe
during auditory and visual oddball paradigms. Electroencephalogr
Clin Neurophysiol 84:269–279.
Ranganath C. 2010. A unified framework for the functional organization
of the medial temporal lobes and the phenomenology of episodic
memory. Hippocampus 20:1263–1290.
Roman R, Brazdil M, Jurak P, Rektor I, Kukleta M. 2005. Intracerebral
P3-like waveforms and the length of the stimulus-response interval in a
visual oddball paradigm. Clin Neurophysiol 116:160–171.
Rosburg T, Trautner P, Ludowig E, Schaller C, Kurthen M, Elger CE,
Boutros NN. 2007. Hippocampal event-related potentials to tone
duration deviance in a passive oddball paradigm in humans. NeuroImage
37: 274–281.
Shin J. 2011. The interrelationship between movement and cognition:
Theta rhythm and the P300 event-related potential. Hippocampus
21:744–752.
Shinba T, Andow Y, Shinozaki T, Ozawa N, Yamamoto K. 1996.
Event-related potentials in the dorsal hippocampus of rats during
an auditory discrimination paradigm. Electroencephalogr Clin
Neurophysiol 100:563–568.
Small SA. 2002. The longitudinal axis of the hippocampal formation:
Its anatomy, circuitry, and role in cognitive function. Rev Neurosci
13:183–194.
Stapleton JM, Halgren E. 1987. Endogenous potentials evoked in
simple cognitive tasks: Depth components and task correlates.
Electroencephalogr Clin Neurophysiol 67:44–52.
Strange BA, Dolan RJ. 2001. Adaptive anterior hippocampal responses
to oddball stimuli. Hippocampus 11:690–698.
Vanderwolf CH. 1969. Hippocampal electrical activity and voluntary
movement in the rat. Electronceph Clin Neurophysiol 26:407–418.
Watrous AJ, Fried I, Ekstrom AD. 2011. Behavioral correlates of
human hippocampal delta and theta oscillations during navigation.
J Neurophysiol 105:1747–1755.
Watrous AJ, Tandon N, Conner CR, Pieters T, Ekstrom AD. 2013.
Frequency-specific network connectivity increases underlie accurate
spatiotemporal memory retrieval. Nat Neurosci 16:349–356.
Wyble BP, Hyman JM, Rossi CA, Hasselmo ME. 2004. Analysis of
theta power in hippocampal EEG during bar pressing and running
behavior in rats during distinct behavioral contexts. Hippocampus
14:662–674.
1344 ROMAN ET AL.
Hippocampus
Příloha č. 3
Kukleta M, Brazdil A, Roman R, Jurak P. Identical event-related potentials to target and
frequent stimuli of visual oddball task recorded by intracerebral electrodes. Clinical
Neurophysiology 2003;114:1292-1297. IF (2003) = 2,485
Identical event-related potentials to target and frequent stimuli
of visual oddball task recorded by intracerebral electrodes
M. Kukletaa,*, M. Bra´zdilb
, R. Romana
, P. Jura´kc
a
Department of Physiology, Medical Faculty, Masaryk University, Brno, Czech Republic
b
First Department of Neurology, Medical Faculty, Masaryk University, Brno, Czech Republic
c
Laboratory of NMR Electronics, Academy of Sciences, Brno, Czech Republic
Accepted 31 March 2003
Abstract
Objective: The shape of visually elicited event-related potentials (ERP) of epileptic patients during their presurgical evaluation with
intracerebral electrodes was investigated in the study.
Methods: Twenty intractable epileptic patients with depth electrodes at several intracranial locations in the frontal, temporal, parietal
lobes, and in the amygdalo-hippocampal complex participated in the study. To evoke the ERP, a standard visual oddball task was used with
target stimuli, and frequent non-habituated and habituated stimuli. The averaged responses of the 3 groups were superimposed and visually
analyzed whether the shape appeared identical or non-identical.
Results: The EEG response to target and frequent stimuli was recorded in 660 intra-cerebral sites. In 88 sites (14 different patients)
localized in the amygdala, parahippocampal gyrus, superior, middle, and inferior temporal gyri, fusiform and lingual gyri, sensorimotor
cortex, prefrontal cortex, hippocampus, and cingulated gyrus, the identical ERPs to target and both groups of frequent stimuli were observed.
In 442 sites located in the above listed structures, and in the basal ganglia and parietal cortex, the shape of the ERP differed from 0.3 to 0.47 s
on after the stimulus. The remaining 130 sites did not yield the task-specific potential change.
Conclusions: The existence of identical ERPs to target and frequent stimuli in the oddball task suggests that a part of mental operations
underlying the brain engagement in this task is not dependent on the way of responding.
q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Event-related potential; Intra-cerebral EEG recording in humans; Oddball task
1. Introduction
In the cognitive brain research, the registration of eventrelated
potentials (ERP) continues to be a basic and everproductive
method. Especially the P3 component, firstly
described by Sutton et al. (1965) and Desmedt et al. (1965),
has been attracting the attention of researchers for decades.
This wave appears in averaged records after rare (target)
external stimuli presented, in a random order, with frequent
ones. It is peaking about 300 ms after an acoustic signal, and
some 100–150 ms later after a visual signal. In the scalp
recordings, its amplitude is largest over the centroparietal
regions for acoustic as well as visual stimulation. The P3
evoked by a visual signal has been analyzed thoroughly
since the 1960s, and available evidence demonstrates its
large dependence on experimental conditions of psychological
character (Donchin et al., 1978; Hillyard and Picton,
1987). This long-latency positive waveform is generally
viewed as reflecting decision or cognitive closure of the
recognition processing, or as the updating of memory for
future actions (Smith et al., 1970; Hillyard et al., 1971;
Hillyard and Picton, 1987; Iragui et al., 1993). Actually, the
P3 is mostly considered as an electrophysiological concomitant
of mental processes linked with the cognitive
elaboration of target signals. Frequent signals are believed
to be unable to produce comparable electrophysiological
changes.
Many studies using scalp electroencephalography
(EEG), intracerebral EEG, and myoelectroencephalography
(MEG) have attempted to localize structures where the P3 is
generated (Basile et al., 1997; ; Baudena et al., 1995;
Bra´zdil, et al., 1999; Goto et al., 1996; Halgren et al., 1995a,b,
1998; Kropotov and Ponomarev, 1991; McCarthy et al.,
Clinical Neurophysiology 114 (2003) 1292–1297
www.elsevier.com/locate/clinph
1388-2457/03/$30.00 q 2003 International Federation of Clinical Neurophysiology. Published by Elsevier Science Ireland Ltd. All rights reserved.
doi:10.1016/S1388-2457(03)00108-1
* Corresponding author. Tel.: þ420-5-42126656; fax: þ420-5-126561.
E-mail address: mkukleta@med.muni.cz (M. Kukleta).
1989; Puce et al., 1991; Rogers et al., 1991; Seeck et al.,
1995; Smith et al., 1990; Tarkka et al., 1995; Tarkka and
Stokic, 1998; Tesche et al., 1996; Yamazaki et al., 2000;
Yamazaki et al., 2001). These studies have shown that the
P3 (and P3 magnetic) is a complicated phenomenon, which
involves many areas of the brain in space and time. The
evidence emerging shows that the P3 reflects parallel and
serial neuronal activity in numerous structures, and that this
activity underlies, very probably, a number of mental
operations necessary for accomplishing the experimental
task. It also becomes clear that the whole set of these
operations has not been delineated unequivocally so far.
In this study, the shape of visually elicited ERP of
epileptic patients with intracerebral electrodes in an oddball
setup has been investigated. Contrary to the common
approach in this method, which consists in differentiating
responses to target and frequent stimuli we were looking for
identical ERP after target and frequent stimuli. The
existence of such responses could be expected when
considering the demands on signal discrimination and
decision-making in the case of frequent stimuli. There is
no reason to believe that these demands are principally
different from those induced by target ones. To confirm this
hypothesis, we chose a simple methodological procedure.
We superimposed and analyzed visually all the ERPs
elicited by target, frequent non-habituated, and frequent
habituated stimuli with the aim to see whether the shape of
these 3 curves appeared identical or non-identical.
2. Methods and materials
2.1. Subjects
Twenty patients (14 males, 6 females; aged 19–47 years;
all with medically intractable epilepsies; 19 right handed,
one left handed) participated in the study. Depth orthogonal
platinum electrodes were implanted to localize the seizure
origin prior to surgical treatment in the frontal, temporal,
and/or parietal lobes using the methodology of Talairach
et al. (1967). In 7 patients, additional diagonal electrodes
were inserted stereotactically into the amygdalo-hippocampal
complex (via the frontal approach, passing through the
basal ganglia in 6 patients, via the occipital approach in one
patient). The electrodes were placed bilaterally in 14
patients and unilaterally in 6 patients. Standard MicroDeep
semi-flexible electrodes (DIXI) with the diameter of
0.8 mm, length of each contact of 2 mm, and inter-contact
intervals of 1.5 mm were used for invasive EEG monitoring.
Contacts at the electrode (5–15) were always numbered
from the medial to lateral sites. Their positions were
indicated in relation to the axes defined by the Talairach
system using the ‘x, z, y’ format where ‘x’ is lateral,
millimeters to midline, positive right hemisphere, ‘z’ is
antero-posterior, millimeters to the AC (anterior commissure)
line, positive anterior, and ‘y’ is vertical, millimeters to the
AC/PC (posterior commissure) line, positive up. The exact
positions of the electrodes and their contacts in the brain
were verified using post-placement magnetic resonance
imaging (MRI) with electrodes in situ. The recordings from
lesional structures and epileptogenic zones were not
included into the analysis. No patient from the group
examined has had bilateral hippocampal sclerosis or
bilateral temporal lobe epilepsy. All the patients had normal
or corrected-to-normal vision. Informed consent was
obtained from each patient prior to the experiment, and
the study received an approval from the Ethical Committee
of Masaryk University.
2.2. Procedure
The patients were seated comfortably in a moderately
lighted room with a monitor screen positioned approximately
100 cm in front of their eyes. During the examination,
they were asked to focus the gaze continuously on the
point in the center of the monitor screen and to respond, as
quickly as possible, to a target stimulus (yellow letter X on
the white background) by pressing a micro-switch button in
the dominant hand and counting the number of these stimuli
in their heads, and to ignore frequent stimuli (yellow letter O
on the white background). Both stimuli were displayed on a
black screen, subtended at the visual angle of 38. Their
duration was 200 ms. The inter-stimulus interval varied
randomly between 2 and 5 s, the ratio of target to frequent
stimuli was 1:5.
2.3. EEG recording
The EEG signal was recorded simultaneously from
various intra-cerebral structures using 64 channel Brain
Quick EEG system (Micromed). All the recordings were
monopolar with respect to a reference electrode placed in all
the cases on the right processus mastoideus. EEGs were
amplified with the bandwidth of 0.1–40 Hz at the sampling
rate of 128 Hz. Further processing was performed with
artifact-free EEG periods (selection was based on visual
inspection of the periods by an experienced person). EEG
periods of 2 s were averaged off-line using the stimulus
onset as the trigger (2500 and þ1500 ms from stimulus
onset). ScopeWin software was used for the signal analysis,
which included up to 44 channels recorded simultaneously.
Responses to frequent stimuli were distributed to a subgroup
of ‘non-habituated responses’ and to a subgroup of
‘habituated responses’, and averaged independently. The
former subgroup gathered responses to the first frequent
stimulus following the target response, the latter one
gathered responses to the sixth, seventh and next (if any)
frequent stimuli following the target response. The mean
value of averaged records was 52 records in the groups of
target responses (median 54, maximum 61, minimum 36),
42 records in the group of frequent ‘non-habituated’
responses (median 44, maximum 53, minimum 29), and
M. Kukleta et al. / Clinical Neurophysiology 114 (2003) 1292–1297 1293
89 records in the group of frequent ‘habituated’ responses
(median 88, maximum 128, minimum 56).
2.4. Data analysis
In total, 660 averaged electrophysiological responses to
target stimuli, 660 averaged responses to non-habituated
frequent stimuli, and 660 averaged responses to habituated
frequent stimuli from different anatomical sites were
evaluated. The number of explored sites in one patient
varied from 14 to 42 sites (median 35 sites). The potential
change after the stimulus has been considered as an ERP if
its amplitude was greater than the twofold of the maximal
potential change seen in the period prior to the stimulus
onset. The decision whether the shape of ERPs to target and
frequent stimuli is identical or non-identical was based on
the visual comparison of superimposed averaged curves
from these 3 conditions. The agreement of two independent
observers was used to decrease a possible subjective factor
in the assessment.
3. Results
Responses from 530 sites (80% of all the investigated
sites) could be considered as event-related potentials
(ERPs). The remaining 20% of sites did not yield taskspecific
potential changes. In 88 sites (14 different patients)
of several structures, the identical ERPs to target and both
groups of frequent stimuli were observed (Table 1). These
sites were situated ipsilaterally in 64% and contralaterally in
36% of cases with regard to the performing hand. As
evident, the occurrence of identical ERPs was more frequent
in the amygdala, parahippocampal gyrus, superior, middle,
and inferior temporal gyrus, in fusiform gyrus and lingual
gyrus. The occurrence of identical responses was less
frequent in the sensorimotor cortex, basal and dorsolateral
prefrontal cortex, hippocampus, and very rare in the
cingulate gyrus. In 72 sites located in the basal ganglia
and parietal cortex, the identical ERP was not found.
Fig. 1 presents 12 cases of identical ERPs induced by
target and frequent stimuli that were selected with the
intention to demonstrate typical ones. The shape of these
potentials varied across different structures as well as within
the same structure. Simple forms, mostly biphasic curves,
were observed frequently in the amygdala. In the other
structures, the shape of identical ERPs was more complex.
In the majority of cases, these potentials included late
components located in the periods, which followed the
execution of movement in responses to a target stimulus.
In another 442 sites of all the investigated structures, the
response to a target stimulus quit the initially identical
course with responses to frequent stimuli, and started its
own course. The instant of these dissociations varied
between 0.30 and 0.47 s. Fig. 2 demonstrates 12 such
cases. In the great proportion of cases, these sites exhibited,
after the dissociation point, task-specific EEG activity even
in response to frequent stimuli (though different from target
responses). The simultaneous occurrence of identical and
non-identical ERPs to target and frequent stimuli in the
same patient and in the same structure was a common
finding. This phenomenon has been observed almost in all
the patients in whom identical responses were demonstrated
(12 out of 14 patients).
The data in Fig. 3 demonstrate 3 regions exhibiting signs
of a local generator. Two of these generators concern the
ERP evoked by a target stimulus only (contacts X0
4, X0
5,
and G1, G2, respectively), the third one could be
considered, at least in its early half, as the generator of
identical ERP (contacts D4, D5).
4. Discussion
The first question to be discussed is the relevance of the
results for understanding the functioning of the healthy brain.
Table 1
The prevalence of identical and non-identical ERPs induced by target and frequent stimuli across different brain structures
Structure Total Identical Non-identical No response
Amygdala 39 (11 patients) 12 (5 patients) 15 (6 patients) 12 (3 patients)
Parahippocampal gyrus 34 (8 patients) 10 (3 patients) 17 (6 patients) 8 (6 patients)
Temporal gyri (superior, middle, inferior) 190 (17 patients) 33 (8 patients) 91 (13 patients) 65 (14 patients)
Fusiform gyrus 15 (8 patients) 5 (2 patients) 9 (6 patients) 1 (1 patient)
Lingual gyrus 5 (1 patient) 3 (1 patient) 2 (1 patient) 0
Sensorimotor cortex 41 (4 patients) 4 (1 patient) 32 (4 patients) 5 (1 patient)
Prefrontal cortex (basal, dorsolateral) 144 (12 patients) 13 (3 patients) 125 (12 patients) 6 (3 patients)
Hippocampus 80 (14 patients) 7 (3 patients) 47 (10 patients) 25 (8 patients)
Cingulate gyrus 40 (11 patients) 1 (1 patient) 34 (10 patients) 5 (1 patient)
Parietal cortex 23 (3 patients) 0 23 (3 patients) 0
Basal ganglia 49 (6 patients) 0 47 (6 patients) 2 (1 patient)
Total, the number of all the explored sites; Identical, the number of sites from which identical ERPs to target and frequent stimuli were derived; Nonidentical,
the number of sites from which non-identical ERPs to target and frequent stimuli were derived; No response, absence of task-specific EEG change
after a stimulus.
M. Kukleta et al. / Clinical Neurophysiology 114 (2003) 1292–12971294
The intracerebral ERP data analyzed in this study as well as
those published previously by various authors were obtained
fromintractable epileptic patientsinthe courseoftheirclinical
presurgical evaluation. Unfortunately, we are always dealing
with the diseased brain tissue in this kind of research.
Considering the presence offunctional and possibly structural
focal lesions in epileptic patients, only EEG signal from
carefully selected extralesional sites can be evaluated. In this
study, recordings from lesional anatomical structures, epileptogenic
zones and sites with obvious epileptiform discharges
were not included in the analysis. Despite the fact it is
impossible to eliminate unequivocally the possible impact of
spreading epileptic activity on the normal function of other
partsofthebrain,wherever theyare located andwhatevertheir
distance from epileptogenic focus is. Another negative
consequence offocal epileptic activity can be a compensatory
shift of physiological functions into distinct cortical locations.
Other limitations of such studies are imposed by ethical
imperatives. In our patients, the locations of electrodes were
strictly determined by diagnostic purposes. Therefore, the
recordings and evaluations of this study can only give
information on the brain areas, which were limited in
number and which were selected unsystematically with
regard to the studied function. For the same reason,
simultaneous or subsequent recordings from the scalp
have not been performed in our patients. Nevertheless,
while keeping in mind all the limitations of the method, its
value for understanding the ERP genesis is currently
unquestionable.
The demonstration of identical intracerebral ERPs to
target and frequent stimuli in oddball task is the main result
of the study presented. This result represents no contradiction
to the consistent findings of different EEG responses to
targets and non-targets. It simply demonstrates that, apart
Fig. 1. Identical averaged ERPs evoked by target (thick line), habituated,
and non-habituated frequent stimuli (thin lines). No of the case, initials of
the patient, location of recording contact, side with respect to the
movement, Talairach’s coordinates: (1) DJ, amygdala, ipsilateral, þ23,
26, 215; (2) FP, amygdala, ipsilateral, þ21, 25, 216; (3) VP, amygdala,
contralateral, 221, 26, 216; (4) KJ, amygdala, contralateral, 223, 26,
216; (5) PM, amygdala, ipsilateral, 223, 25, þ15; (6) KJ, fusiform
gyrus, contralateral, 229, 225, 216; (7) KP, superior temporal gyrus,
ipsilateral, þ60.5, 233, þ17; (8) KZ, supplementary motor area,
ipsilateral, þ13, 212, þ52; (9) DM, parahippocampal gyrus, ipsilateral,
þ18, 243, 0; (10) MP, lingual gyrus, ipsilateral, þ23, 245, 21; (11) KZ,
dorsolateral prefrontal cortex, ipsilateral, þ54, þ4, þ31; (12) MP, middle
temporal gyrus, ipsilateral, 264, 244, þ2.
Fig. 2. Non-identical average ERPs evoked by target (thick line),
habituated, and non-habituated frequent stimuli (thin lines). No of the
case, initials of the patient, location of recording contact, side with respect
to the movement, Talairach’s coordinates: (1) SM, putamen, ipsilateral,
þ26, 27, þ8; (2) HZ, fusiform gyrus, ipsilateral, þ34, 242, 25; (3) PM,
middle frontal gyrus, ipsilateral, 214, þ44, 28; (4) FP, parahippocampal
gyrus, ipsilateral, þ18, 234, 23; (5) DM, amygdala, contralateral, 225,
25, 217; (6) KZ, suplementary motor area, contralateral , 21, 29, þ51;
(7) DM, hippocampus posterior, contralateral, 225, 227, 28; (8) SM,
inferior temporal gyrus, ipsilateral, þ58, 226, 213; (9) PM, anterior
cingulated gyrus, contralateral, þ4, þ22, þ32; (10) KZ, sensorimotor
cortex, contralateral, 228.5, 29, 251; (11) JI, parahippocampal gyrus,
ipsilateral, þ24, 213, 224; (12) DJ, parahippocampal gyrus, contralateral,
223, 26, 24.5.
M. Kukleta et al. / Clinical Neurophysiology 114 (2003) 1292–1297 1295
from sites where target stimuli are processed specifically,
there can be found other sites, in which the processing of
targets and frequents is going identically. We have not
found, in the pertinent literature, a study looking for
identical ERP after target and non-target stimuli with
intracranial electrodes. The identical ERP had predominantly
a form of slow biphasic potential change that
extinguished about 1 s after the signal. The shape of
identical ERPs varied across the structures explored and,
sometimes, even across different contacts located in the
same structure. The unstable form of ERPs was a frequent
finding, and it reflected, very probably, the complexity of
factors in space and time that determined the characteristics
of electrical field in the surroundings of recording contacts.
The orientation of functional dipoles, which generated these
fields, their power, form, and number as well as the tissue
conductivity – these are some possibilities of determining
factors. In such a situation, the comparison of records
obtained from the same site under various conditions can
only provide reliable results. In the present study, the
creation of two subgroups of frequent responses (nonhabituated
and habituated subgroups) was inspired by this
methodological requirement. The increased number of
curves superimposed in evaluating their shape and the
resulting increase of reliability of this procedure was the
only reason of this subdivision.
As demonstrated, the identical and non-identical
responses occurred frequently in the same patient and in
the same structure. Identical responses were less frequent;
they represented 17% of identified ERPs only. Another
aspect of the results necessitates a comment. Amongst 88
identical ERPs recorded in different brain structures, the
generator was demonstrated on one occasion only (Fig. 3,
contacts D4, D5). All the other evaluated identical ERPs
have been generated in the sites more or less remote from
the recording contacts. This fact has considerably limited
any attempt of localizing the structures responsible for their
generation.
The study has not yielded sufficient data for the
interpretation of behavioral significance of identical ERP.
Nevertheless, it is clear that the demonstration of identical
ERP further corroborate the complicated image of the P3
phenomenon emerging from relevant EEG studies. The
demonstration of simultaneous occurrence of identical and
non-identical ERPs to target and frequent stimuli in an
oddball task allows differentiating between responsedependent
and response-independent intra-cerebral ERPs.
This suggests the existence of response-dependent and
response-independent mental operations that, at least in
some cases (see Fig. 3, contacts G1 and G2), are organized
in a parallel fashion. This reasoning is not supported directly
by the data presented; nevertheless, it has an inspiring
quality for orienting further research in this field.
Acknowledgements
We would like to thank Dr M. Dufek for his valuable
assistance. The results were presented at the 37th Conference
on Higher Nervous Functions, Olomouc, September
2001.
References
Baudena P, Halgren E, Heit G, Clarke JM. Intracerebral potentials to rare
target and distractor auditory and visual stimuli. III. Frontal cortex.
Electroenceph clin Neurophysiol 1995;94:251–64.
Basile LF, Rogers RL, Simos PG, Papanicolaou AC. Magnetoencephalographic
evidence for common sources of long latency fields to rare
target and rare novel visual stimuli. Int J Psychophysiol 1997;25:
123–37.
Bra´zdil M, Rektor I, Dufek M, Daniel P, Jura´k P, Kuba R. The role of
Fig. 3. The averaged ERPs evoked by target (thick line), habituated, and
non-habituated frequent stimuli (thin lines), recorded from neighboring
electrode contacts exhibiting phase reversal and steep voltage gradients.
Left half: patient DJ, plots in parahippocampal gyrus (X0
3, 223, 26,
24.5), amygdala (X0
4, 223, 25, 221), amygdala (X0
5, 223, 24,
217.5), amygdala (X0
6, 223, 23, 214), amygdala (X0
7, 223, 22,
210.5), white matter (X0
8, 223, 21, 27), putamen (X0
9, 223, 0, 23.5);
side contralateral with respect to the movement. Right half, above: patient
SM, plots in cingulated gyrus (G1, þ3, 235, þ28; G2, þ6.5, 235, þ28;
G3, þ10, 235, þ28; G4, þ13.5, 235, þ28; G5, þ17, 235, þ28); side
ipsilateral with respect to the movement. Right half, below: patient KP,
superior temporal gyrus (D2, þ53.5, 233, þ17; D3, þ57, 233, þ17; D4,
þ60.5, 233, þ17; D5, þ64, 233, þ17); side ipsilateral with respect to
the movement.
M. Kukleta et al. / Clinical Neurophysiology 114 (2003) 1292–12971296
frontal and temporal lobes in visual discrimination task-depth ERP
studies. Neurophysiol Clin 1999;29:339–50.
Desmedt JE, Debecker J, Manil J. Mise en evidence d’un signe e´lectrique
cerebral associe´ a la detection par le sujet d’un stimulus sensoriel
tactile. Bull Acad R Med Belg 1965;5:887–936.
Donchin E, Ritter W, McCallum C. Cognitive psychophysiology: the
endogenous components of the ERP. In: Callaway E, Tueting P, Koslow
S, editors. Brain event-related potentials in man. New York, NY:
Academic Press; 1978. p. 349–411.
Goto Y, Brigell MG, Parmeggiani L. Dipole modeling of the visual evoked
P300. J Psychosom Res 1996;41:71–9.
Halgren E, Baudena P, Clarke JM, Heit G, Lie´geois C, Chauvel P, et al.
Intracerebral potentials to rare target and distractor auditory and visual
stimuli. I. Superior temporal plane and parietal lobe. Electroenceph clin
Neurophysiol 1995a;94:191–220.
Halgren E, Baudena P, Clarke JM, Heit G, Marinkovic K, Deveaux B, et al.
Intracerebral potentials to rare target and distractor auditory and visual
stimuli. II. Medial, lateral and posterior temporal lobe. Electroenceph
clin Neurophysiol 1995b;94:229–50.
Halgren E, Marinkovic K, Chauvel P. Generators of the late cognitive
potentials in auditory and visual oddball tasks. Electroenceph clin
Neurophysiol 1998;106:156–64.
Hillyard SA, Picton TW. Electrophysiology of cognition. In: Plum F,
editor. Higher functions of the brain, Part 2. Handbook of Physiology,
Section 1 (The nervous syste´m 5), Washington, DC: American
Physiological Society; 1987. p. 591–684.
Hillyard SA, Squires KC, Bauer JW, Lindsay PH. Evoked potential
correlates of auditory signal detection. Science 1971;172:1357–60.
Iragui VJ, Kutas M, Mitchiner MR, Hillyard SA. Effect of aging on eventrelated
brain potentials and reaction times in an auditory oddball task.
Psychophysiology 1993;30:10–22.
Kropotov JD, Ponomarev VA. Subcortical neuronal correlates of
component P300 in man. Electroenceph clin Neurophysiol 1991;78:
40–9.
McCarthy G, Wood CC, Williamson PD, Spencer DD. Task-dependent
field potentials in human hippocampal formation. J Neurosci 1989;9:
4253–60.
Puce A, Andrewes DG, Berkovic SF, Bladin PF. Visual recognition
memory. Neurophysiological evidence for the role of temporal white
matter in man. Brain 1991;114:1647–66.
Rogers RL, Baumann SB, Papanicolaou AC, Bourbon TW, Alagarsamy S,
Eisenberg HM. Localization of the P3 sources using magnetoencephalography
and magnetic resonance imaging. Electroenceph clin
Neurophysiol 1991;79:308–21.
Seeck M, Schomer D, Mainwaring N, Ives J, Dubuisson D, Blume H, et al.
Selectively distributed processing of visual object recognition in the
temporal and frontal lobes of the human brain. Ann Neurol 1995;37:
538–45.
Smith DBD, Donchin E, Cohen L, Starr A. Auditory evoked potentials in
man during selective binaural listening. Electroenceph clin Neurophysiol
1970;28:146–52.
Smith ME, Halgren E, Sokolik M, Baudena P, Musolino Liegeois-Chauvel
C, Chauvel P. The intracranial topography of the P3 event-related
potential elicited during auditory oddball. Electroenceph clin Neurophysiol
1990;76:235–48.
Sutton S, Baren M, Zubin J, John ER. Evoked potentials correlates of
stimulus uncertainty. Science 1965;150:1187–8.
Talairach J, Szikla G, Tournoux P, Prosalentis A, Bordas-Ferrer M, Covello
J, et al. Atlas d’Anatomie Stereotaxique du Telencephale. Paris:
Masson; 1967.
Tarkka IM, Stokic DS. Source localization of P300 from oddball, single
stimulus, and omitted-stimulus paradigmas. Brain Topogr 1998;11:
141–51.
Tarkka IM, Stokic DS, Basile LF, Papanicolaou AC. Electric source
localization of the auditory P300 agrees with magnetic source
localization. Electroenceph clin Neurophysiol 1995;96:538–45.
Tesche CD, Karhu J, Tissari SO. Non-invasive detection of neuronal
population activity in human hippocampus. Brain Res Cogn Brain Res
1996;4:39–47.
Yamazaki T, Kamijo K, Kenmochi A, Fukuzumi S, Kiyuna T, Takaki Y,
Kuroiwa Y. Multiple equivalent current dipole source localization of
visual event-related potentials during oddball paradigm with motor
response. Brain Topogr Spring 2000;12:159–75.
Yamazaki T, Kamijo K, Kiyuna T, Takaki Y, Kuroiwa Y. Multiple dipole
analysis of visual event-related potentials during oddball paradigm with
silent counting. Brain Topogr Spring 2001;13:161–8.
M. Kukleta et al. / Clinical Neurophysiology 114 (2003) 1292–1297 1297
Příloha č. 4
Roman R, Chládek J, Brázdil M, Jurák P, Rektor I, Kukleta M. Changes of oscillatory activity
in a visual oddball task (sEEG study). Homeostasis in Health and Disease 2006;44(4):169-171.
Příloha č. 5
Brazdil M, Roman R, Daniel P, Rektor I. Intracerebral error-related negativity in a simple
Go/NoGo task. Journal of Psychophysiology 2005;19:244-255. IF (2005) = 0,968
Used by permission from the Journal of Psychophysiology 2005, Vol. 19(4):244-255
©2005 Hogrefe Publishing (formerly Hogrefe & Huber Publishers) www.hogrefe.com
DOI 10.1027/0269-8803.19.4.244
M. Brázdil et al.: Neural Sources of Error-Related NegativityJournal of Psychophysiology 2005; Vol. 19(4):244–255Hogrefe&HuberPublishers
Intracerebral Error-Related
Negativity in a
Simple Go/NoGo Task
Milan Brázdil1
, Robert Roman2
, Pavel Daniel1
, and Ivan Rektor1
1
First Department of Neurology, St. Anne’s University Hospital, Brno, Czech Republic
2
Department of Physiology, Masaryk University, Brno, Czech Republic
Abstract. Performance monitoring represents a critical executive function of the human brain. In an effort to identify its anatomical
and physiological aspects, a negative component of event-related potentials (ERPs), which occurs only on incorrect trials,
has been used in the extensive investigation of error processing. This component has been termed “error-negativity” (Ne) or
error-related negativity (ERN) and has been interpreted as a correlate of error detection. The aim of the present intracerebral ERP
study was to contribute knowledge of the sources of the Ne/ERN, with a particular focus on the involvement of a frontomedian
wall (FMW) in the genesis of this negativity. Seven patients with intractable epilepsy participated in the study. Depth electrodes
were implanted to localize the seizure origin prior to surgical treatment. A total of 574 sites in the frontal, temporal, and parietal
lobes were investigated. A simple Go/NoGo task was performed and EEG epochs with correct and erroneous motor responses
were averaged independently using the response as the trigger.
Ne/ERN was generated in multiple cortical structures, with the most consistent involvement being that of the FMW structures.
Ne/ERN generators were revealed there in both the rostral and caudal anterior cingulate cortex (ACC), but also in the pre-SMA
and in the parts of the medial frontal gyrus adjacent to the ACC. Different timing of activations between the rostral and caudal
anterior cingulate Ne/ERN sources was observed in this study. Other neural sources of the Ne/ERN were found in the dorsolateral
prefrontal cortex, in the orbitofrontal cortex, in the lateral temporal neocortex, and in one isolated case in the supramarginal gyrus.
Our findings support the key role of the FMW in the genesis of Ne/ERN. At the same time, our findings suggest a different
functional significance for the rostral and caudal ACC involvement in error processing. In addition to the FMW, the other
prefrontal cortical sites, the lateral temporal neocortex, and the supramarginal gyrus seem to represent integral components of
the brain’s error monitoring system.
Keywords: error processing, event-related potentials, intracerebral recordings, Go/NoGo task
In the early 1990s, two groups independently discovered
a negative component of event-related potentials
(ERPs), which occurred only on incorrect trials (Falkenstein,
Hohnsbein, & Hoormann, 1991; Gehring et al.,
1993). This negativity shows a frontocentral maximum
in scalp recordings, peaking about 50–100 ms after the
erroneous motor response. It has been termed “error negativity”
(Ne) or “error-related negativity” (ERN) and has
generally been interpreted as a correlate of error detection,
or alternatively as a correlate of response checking
itself (Coles, Scheffers, & Holroyd, 2000; Falkenstein et
al., 2000; Vidal, Hasbroucq, & Bonnet, 2000). It is noteworthy,
however, that many individuals also exhibit a
smaller response-locked frontocentral negativity after
the execution of correct responses. This component is
known as the “correct response negativity” (CRN); its
amplitude is usually much smaller than that of the
Ne/ERN (Falkenstein et al., 2000; Ford, 1999; Vidal et
al., 2000). Ne/ERN is an online index of performance
monitoring that may be independent of response conflict
because it is observed in simple choice-reaction tasks
without response competition (Falkenstein et al., 2000).
However, it is possible that Ne/ERN can also signal the
detection of conflict between the executed error and the
correct response that becomes activated through continued
processing of the stimulus (Botvinick et al., 2001;
Yeung, Cohen, & Botvinick, 2004). Thus, despite its
complexity, Ne/ERN can be favorably used in the study
of the neural substrates of error processing.
An increasingly animated issue in the research of performance
monitoring concerns the site where the specific
error-detection system is implemented in the brain.
Earlier dipole source localization analyses suggested a
local genesis of the Ne/ERN within structures of the
DOI 10.1027/0269-8803.19.4.244
© 2005 Federation of European Psychophysiology Societies
Journal of Psychophysiology 2005; Vol. 19(4):244–255 Hogrefe & Huber Publishers
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
frontomedian wall (FMW) – most probably in the anterior
cingulate cortex (ACC) or presupplementary motor
area (pre-SMA) (Dehaene, Posner, & Tucker, 1994; Holroyd,
Dien, & Coles, 1998; Luu, Flaisch, & Tucker,
2000; Miltner et al., 1998; Ridderinkhof et al., 2004; Van
Veen & Carter, 2002a, b). It must be noted that source
modeling of electrical components of the ERPs is limited
in its spatial resolution because there is no unique solution
to the inverse problem. Nevertheless, recent studies
using functional magnetic resonance imaging (fMRI)
have also revealed significant activation of the ACC during
erroneous trials (Braver et al., 2001; Carter et al.,
1998; Fiehler, Ullsperger, & von Cramon, 2004; Garavan
et al., 2002, 2003; Kiehl, Liddle, & Hopfinger, 2000;
Menon et al., 2001; Ullsperger & von Cramon, 2001,
2004). Less frequently, additional activations were also
observed in the pre-SMA, the insular cortex, the inferior
parietal lobule, the supramarginal gyrus, the orbitofrontal
cortex, the posterior cingulate, and the thalamus
(Garavan, Hester, & Fassbender, 2004; Menon et al.,
2001; Ullsperger & von Cramon, 2001). However, the
theory that the Ne/ERN originates in the ACC, or within
other brain sites with significant hemodynamic response
to errors, has not yet been directly confirmed.
Intracerebral ERP recordings, performed in epilepsy
surgery candidates during preoperative invasive longterm
video-EEG monitoring, represent a good opportunity
for directly defining the generators of cognitive
evoked potentials. Error-related potentials elicited by incorrect
responses to frequent stimuli in a visual oddball
task were recently analyzed using standard methodology
(Brázdil et al., 2002). The results stressed the role of
multiple cortical structures in the genesis of observed
Ne/ERN. In addition to the anterior cingulate, the mesiotemporal,
and some prefrontal cortical sites repeatedly
generated Ne/ERN-like potentials during the performed
task. In terms of experimental tasks, this analysis differed
significantly from previous ERP and fMRI studies
on error processing, so mutual comparisons of the results
remain somewhat problematic. The aim of the present
intracerebral ERPstudy was to contribute more accurately
to the knowledge of cerebral generators of the
Ne/ERN by using a standard Go/NoGo task. Special attention
was devoted to the involvement of the FMW in
the genesis of Ne/ERN.
Methods
Subjects
Seven patients (six males and one female) ranging in age
from 23 to 35 years (with an average age of 29.4 years),
all with medically intractable epilepsies, participated in
the study (Table 1). The epileptogenic zone was found in
the mesiotemporal regions in four subjects (three left,
one right), within the frontal lobe structures in two subjects
(one ACC right, one frontopolar cortex left); and in
the parietal lobe in one subject (postcentral region left).
A comprehensive neuropsychological examination excluded
severe cognitive disturbances and dementia in
each patient. Depth electrodes were implanted to localize
the seizure origin prior to surgical treatment. Each patient
received 5–11 orthogonal multicontact electrodes
using the methodology of Talairach et al. (1967). A total
of 574 sites in the frontal, temporal, and parietal lobes
were investigated. The number of sites per patient varied
from 59 to 114. Standard MicroDeep semiflexible electrodes
(DIXI Medical, Besançon, France) with diameters
of 0.8 mm, contact lengths of 2 mm, and intercontact intervals
of 1.5 mm were used for invasive EEG monitoring.
Contacts at the electrodes (5–15) were always numbered
from the medial to the lateral side. Their positions
were indicated in relation to the axes defined by the Talairach
system using the “x,y,z” format where “x” = lateral,
mm to midline, positive right hemisphere; “y” =
anteroposterior, mm to the AC line, positive anterior; and
“z” = vertical, mm to the AC-PC (anterior commissureposterior
commissure) line, positive up. The exact positions
of the electrodes and their contacts in the brain were
verified using postplacement magnetic resonance imaging
(MRI) with electrodes in situ. The recordings from
Table 1. Patient characteristics.
Pt. No Sex Age (years) Dominant hand Implanted sites* Recording sites Error rate (%) Reaction time
(ms)
1 M 31 R LF, RF 84 35.6 381 ± 82.5
2 M 29 R LTFP 68 39.6 372 ± 71.7
3 M 31 L LTF, RT 84 79.7 286 ± 42.3
4 M 35 R LT, RT 65 36.7 316 ± 36.5
5 M 23 R LTF 59 62.1 337 ± 41.3
6 F 23 R RTF 114 28.3 410 ± 29.0
7 M 34 R LTF, RTFP 100 40.0 422 ± 40.5
*T = temporal; F = frontal; P = parietal; R = right; L = left.
M. Brázdil et al.: Neural Sources of Error-Related Negativity 245
Hogrefe & Huber Publishers Journal of Psychophysiology 2005; Vol. 19(4):244–255
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
lesional anatomical structures and epileptogenic zones
were not included in the analysis. All subjects had normal
or corrected-to-normal vision. Informed consent
was obtained from each subject prior to the experiment
and the study received the approval of the Ethics Committee
of Masaryk University.
Procedure
The experiment was conducted on each patient in the
morning hours during a long-term video-EEG monitoring.
Subjects were seated comfortably in a moderately
lighted room with a monitor screen positioned approximately
60 cm in front of their eyes. During the examination,
they were requested to focus their eyes continuously
on the small fixation point in the center of the screen,
and to minimize blinking. Two types of visual stimuli –
clearly visible, yellow uppercase letters “X” (0.80 probability;
approx. 250 trials) and “K” (0.20 probability; 60
trials) – were presented in the center of the screen. With
the exception that two “K”s were never presented sequentially,
the order of “X”s and “K”s was completely
random. The stimuli were approximately 3 × 5 visual
degrees and were presented for 240 ms on a black background.
The interstimulus interval varied randomly between
650, 1650, and 2650 ms. Participants were instructed
to respond as quickly and as accurately as possible
with their dominant index finger on a computer
keyboard every time the “X” appeared, and to not respond
to the “K.” Reaction time and accuracy were
equally stressed. In this task, subjects relatively often
made erroneous responses (false alarms) to the “K”s
(i.e., incorrectly hit the keyboard). Prior to the recording,
each participant performed a block of 10 practice trials
twice to ensure that they understood the instructions. After
the recording session, reaction times for false alarms
and error rates were computed for each subject.
EEG Recording
The EEG signal was simultaneously recorded from various
intracerebral structures, and from standard Fz, Cz,
and Pz scalp electrodes, using the 96 channel Brain
Scope EEG system (M&I; Patients 1–5) or the 128 channel
TrueScan EEG system (Deymed Diagnostic; Patients
6–7). All recordings were monopolar, with a linked earlobe
reference. All impedances were less than 5 kΩ. Eye
movements were recorded from a cathode placed 1 cm
lateral and 1 cm above the canthus of the left eye, and
from an anode 1 cm lateral and 1 cm below the canthus
of the right eye. In each subject, simultaneous bipolar
surface EMG recordings from the dominant finger interossei
muscle were obtained as a control of the subject’s
motor response. EEGs were amplified with a complete
bandwidth at a sampling rate of 256 Hz. Occasional
EEG artifacts were rejected manually during off-line
analysis and further processing was performed with artifact-free
EEG periods. For each patient, EEGs were analyzed
using ScopeWin software. EEG data was digitally
filtered with a zero-phase shift 10 Hz low-pass filter, and
a zero-phase shift 2 Hz high-pass filter. EEG epochs of
600 ms were averaged off-line using the motor response
as the trigger (response-triggered averages [RTA]; 300
and +300 ms from the response). EEG epochs with correct
and incorrect motor reactions were averaged separately
for each subject. All averages were baseline corrected
to a 50 ms period beginning 200 ms before subject
response. The main ERP components in the latency
range of –100 ms – +200 ms were independently identified
by visual inspection by two authors, and quantified
by latency measures. The most valuable discoveries
were the less frequent intracerebral findings of polarity
reversals (identical potentials with opposite polarity in
recordings from two adjacent contacts of the same intracerebral
electrode) and the steep voltage gradients of the
recorded potentials that uniquely proved the focal origin
of the waveform (Vaughan et al., 1986). To prevent confusing
the intracerebral P3-like potentials with the errorrelated
potentials (see Results), the stimulus-triggered
ERPs (–100 and +500 ms from the stimulus) were subsequently
computed for error and correct trials after rare
“K”s in the brain regions, where the Ne/ERN generators
were clearly revealed in the response-triggered averages.
For each anatomical structure investigated by depth electrode,
the measurement values from the recording channel
(single electrode contact) with the most prominent
potential responses were chosen for further analysis.
Two-way MANOVAs (electrode × response type)
were used for repeated measurements in the experiment
to test the effect of the response type (correct hits vs. false
alarms) at the three scalp electrode sites (Fz, Cz, and Pz)
on the negative peak voltage (measured to a baseline)
and latency. Subsequently a paired t-test was used for
evaluating the effect of the response type on the ERP
latencies and amplitudes at each scalp electrode separately.
Statistics were obtained by using the routines included
in the Statistica program (StatSoft).
Results
Satisfactory cooperation of all subjects was observed
during the experiment. The mean false-alarm rate in our
investigated subjects was 46% (SD = 18.17). The mean
reaction time for false alarms was 361 ms (SD = 49.8).
246 M. Brázdil et al.: Neural Sources of Error-Related Negativity
Journal of Psychophysiology 2005; Vol. 19(4):244–255 Hogrefe & Huber Publishers
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
In scalp recordings, Ne/ERN was clearly expressed
after erroneous responses in most subjects and generally
it was maximal at the Fz electrode (mean amplitude
–7.6 μV; mean latency 84 ms). The grand averages of
scalp ERPs for response-locked correct trials and response-locked
error trials across the investigated subjects
are presented in Figure 1. Mean latencies and amplitudes
of scalp ERPs for error trials (Ne/ERN) and for
correct trials (CRN) can be found in Table 2. The response
type had a significant effect on the ERN/CRN
amplitude, F(1,30) = 8.89, p = .005, with higher values
for the false alarm than for the correct hit condition. A
strongly significant effect of the response type on the
ERN/CRN latency was also observed, F(1,30) = 14.38,
p = .0007. In contrast, the electrode site did not have a
significant effect on the negativity amplitudes, F(2,30) =
1.15, p = .330, or their latencies, F(2,30) = 0.26, p = .772.
No significant effect on ERN/CRN amplitude and latency
was found in reference to interactions between the
recording site and response type, F(2,30) = 0.91, p =
Table 2. Mean Ne/ERN and CRN amplitude and latency values at
Fz, Cz, and Pz.
Ne/ERN CRN
Mean SD Mean SD
Amplitudes Fz –7.6 3.20 –3.0 1.52
Cz –6.0 4.17 –1.6 3.59
Pz –3.9 3.53 –2.7 3.71
Latencies Fz 84.3 35.10 29.4 39.38
Cz 74.2 46.61 45.7 52.08
Pz 110.7 37.99 27.8 49.22
Figure 1. Grand averages of the RTA for correct responses (thin
lines) and error trials (thick lines) at the scalp electrodes (Fz, Cz,
Pz) and EOG. R = motor response.
Figure 2. Response-triggered averaged intracerebral
and scalp ERPs for correct (thin
lines) and error (thick lines) trials. Averaged
responses to errors reveal Ne/ERNlike
potential without considerable voltage
variation in a number of adjacent electrode
contacts within the left temporal lobe
(B’7–12; approximate coordinates in the
Talairach axes of x = –41 – –61 mm, y =
–22 mm, z = –9 mm) = far field (on the
left). On the right, steep voltage gradients
in adjacent contacts O’2–3 indicate focal
origin of Ne/ERN in the left-sided medial
frontal gyrus (Talairach’s coordinates, –7,
40, –10). Subject no. 5.
M. Brázdil et al.: Neural Sources of Error-Related Negativity 247
Hogrefe & Huber Publishers Journal of Psychophysiology 2005; Vol. 19(4):244–255
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
.413; F(2,30) = 1.15, p = .325. The effect of the response
type on the Ne/ERN amplitude was at its greatest at Fz
(t = –2.77; df = 5; p = .039), lower at Cz (t = –2.21; df =
5; p = .078), and absent at Pz (t = –1.04; df = 5; p = .344).
Intracerebral recordings in a number of cortical locations
revealed prominent Ne/ERN-like potentials (with
RTA-latency in the range of –100 ms – +200 ms) after
false alarms that were not present after correct hits. These
potentials occurred either in a number of adjacent electrode
contacts without considerable voltage variation
(this finding proves a far field of the recorded potential
– spreading from a more or less distant neural source; see
Figure 2), or restrictively with a phase reversal or steep
voltage gradients in adjacent electrode contacts. This
finding proves the focal origin of the waveform (neural
source; see Figures 2 and 3). The number of negative
findings (i.e., no evident difference in averages on correct
and incorrect trials, classified as “absent Ne/ERN”),
far fields, and generators within the investigated brain
regions for all participating subjects is given in Table 3.
Despite extensive investigation by means of depth electrodes,
we never found a focal origin of Ne/ERN-like
potentials within the amygdala, or the superior temporal,
fusiform, or precentral gyri. Far fields of the recorded
distinct ERP to false alarms were frequently observed in
middle temporal gyrus and amygdala. The most interesting
finding was that the focal origin of the ERPs to false
alarms were proven in a number of brain sites – the hippocampus,
the ACC, the medial frontal gyrus, the middle
temporal gyrus, the middle frontal gyrus, the orbito-frontal
cortex, the pre-SMA, and the inferior frontal, inferior
temporal, and supramarginal gyri. As the rarely-presented
“K” stimuli in the performed task were hypothetically
potent enough to activate the neural populations involved
in the genesis of P3 potential (which would not
be activated by the frequent “X” stimuli), the stimulusFigure
3. Intracerebral Ne/ERN with a character indicating its
local source. Recordings from the left-sided rostral cingulate motor
area (F’1–2: Talairach’s coordinates, –5, 15, 40). Simultaneous
recordings from scalp Fz, Cz, and Pz electrodes, EOG and EMG.
Subject no. 3.
Table 3. Summary of the intracerebral findings for investigated
anatomical sites. Number of negative findings, far fields, and generators
of Ne/ERN-like potentials as revealed by RTA technique.
Anatomical site (Brodmann area) Negative
finding
Far
field
Focal
origin
Anterior cingulate cortex (32) 2 4 3
Anterior cingulate cortex (24) 3 2 2
Medial frontal gyrus (8,9,10) 3 1 5
Pre-SMA (6) 1 1 2
Orbito-frontal cortex (11) 1 2 3
Middle frontal gyrus (6, 8, 9, 10, 46) 8 2 3
Inferior frontal gyrus (45, 46, 47) 3 3 2
Precentral gyrus (4, 6) 2 2 0
Amygdala 2 8 0
Hippocampus 1 1 11
Fusiform gyrus (36) 1 0 0
Superior temporal gyrus (22, 38, 42) 7 2 0
Middle temporal gyrus (21) 9 13 5
Inferior temporal gyrus (20) 0 0 1
Angular gyrus (39) 1 0 0
Supramarginal gyrus (40) 0 0 1
248 M. Brázdil et al.: Neural Sources of Error-Related Negativity
Journal of Psychophysiology 2005; Vol. 19(4):244–255 Hogrefe & Huber Publishers
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
Figure 4. RTAs (for correct and error trials)
and STAs (for erroneously hit and correctly
ignored “K”s) for the intracerebral
(right-sided hippocampus – B1–3: 27–36,
–19, –12) and scalp recordings. Intrahippocampal
ERPs of analogous configuration
to false alarms and response inhibition
in STAs dispute the direct relationship of
the hippocampal activation to the error-detection
in the task used. S = stimulus presentation.
Typical NoGo N2 potential after
correctly ignored trials is indicated in scalp
STAs. Subject no. 4.
Table 4. List of Ne/ERN generators with their approximate coordinates in the Talairach axes and individual latencies.
Anatomical site Brodmann area Subject no. Talairach’s coordinates (x,y,z) Ne/ERN latency (ms)
ACC BA 32 3 –5, 15, 40 56
BA 32 1 –7, 20, 30 0
BA 32 1 –5, 34, –5 20
BA 24 1 –2, 31, 14 0
Medial frontal gyrus BA 10 5 –7, 40, –10 45
BA 9 2 –5, 38, 32 70
BA 10 7 3, 41, –10 48
BA 10 7 –4,45, –12 91
Pre-SMA BA 6 2 –4, –5, 53 70
BA 6 6 3, 2, 56 82
Orbito-frontal cortex BA 11 5 –25, 40, –10 22
BA 11 7 –37, 45, –12 28
Middle frontal gyrus BA 8 3 –48, 15, 42 59
BA 46 6 45, 34, 21 20
BA 6 6 30, 2, 56 37
Inferior frontal gyrus BA 46 6 50, 40, 11 37
BA 45 7 53, 24, 19 51
Middle temporal gyrus BA 21 3 –61, –12, –12 14
BA 21 3 –61, –24, –9 50
BA 21 7 53, –19, –9 162
Inferior temporal gyrus BA 20 6 45, –27, –11 –9
Supramarginal gyrus BA 40 7 54, –48, 37 –37
M. Brázdil et al.: Neural Sources of Error-Related Negativity 249
Hogrefe & Huber Publishers Journal of Psychophysiology 2005; Vol. 19(4):244–255
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
triggered ERPs were further computed for error and correct
responses after the presentation of the “K” stimuli to
differentiate the effects of “rare events” (as errors could
only occur on rare trials). This analysis was performed
in brain regions with proof of Ne/ERN-like potential
generators as detected in response-triggered averages.
Analogous ERPs in error and correct responses to “K”s
were repeatedly found in different brain regions. A typical
situation was observed in the intrahippocampal recordings,
where stimulus-triggered averages showed
identical waveforms after both erroneously hit and correctly
ignored “K”s in all the sites with generators of
Ne/ERN-like potentials (Figure 4). Similarly, several
other neural sources of Ne/ERN-like potentials within
the middle temporal gyrus, the orbito-frontal cortex, and
some mesiofrontal sites (Brodmann areas 24 and 10)
were proven to be involved in the processing of rare stimuli,
and not exclusively in specific error processing.
However this behavior was only observed in some recordings
from the described brain structures, and was not
typical for the extrahippocampal structures. Excluding
sources of these pseudo-ERN potentials, we obtained the
list of definite Ne/ERN generators presented in this study
(Table 4). In all of these sites there were obvious differences
in the waveforms obtained from stimulus-triggered
averaging (false alarms after “K”s vs. correct response
inhibition after “K”s). At the same time, RTAs
showed Ne/ERN after error hits with a character indicating
focal origin (see above). It is clear that true Ne/ERN
was generated in multiple cortical structures, with inFigure
5. Schematic map of Ne/ERN findings on the mesial aspect
of the frontal lobe (group data).
Figure 6. A time delay between intracerebral Ne/ERNs generated within different ACC sub-areas (response-triggered averages from
cingulate motor area – CMA: –7, 20, 30 – Ne/ERN latency = 0 ms; BA 24: –2, 31, 14 – Ne/ERN latency = 0 ms; rostral ACC: –5, 34,
–5 – Ne/ERN latency = +20 ms). The position of the recording sites (right) is precisely indicated on the subject’s MRI scan (verified
using postplacement MRI). Subject no. 1. FF – far field (spreading of the Ne/ERN activity).
250 M. Brázdil et al.: Neural Sources of Error-Related Negativity
Journal of Psychophysiology 2005; Vol. 19(4):244–255 Hogrefe & Huber Publishers
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
volvement of the FMW structures being the most consistent.
Ne/ERN generators were revealed there at 10 locations:
most frequently in the rostral and caudal ACC, in
the adjacent medial frontal gyrus (mainly within the ventral
bank of the paracingulate sulcus, in which cytoarchitectonic
and functional properties can be identical or
very similar to those of the paralimbic cortex – BAs 32
and 24c; Paus, 2001), but also in the pre-SMA. A schematic
map of Ne/ERN findings on the mesial aspect of
the frontal lobe is given in Figure 5. A mutual comparison
of Ne/ERN latencies among several generators within
the FMW was possible in one subject (No. 1). This
intraindividual analysis detected at least two separate
Ne/ERN sources in the left-sided FMW. Clearly different
Ne/ERN latencies were found in the left-sided FMW,
with a time delay of 20 ms in the rostral anterior cingulate
in comparison to the latency of Ne/ERN generated in a
more caudal ACC (Figure 6). Other neural sources of
Ne/ERN were proven in the dorsolateral prefrontal cortex,
in the orbito-frontal cortex, in the lateral temporal
neocortex, and in one isolated case within the supramarginal
gyrus. The latency values of intracerebral Ne/ERNs
at the sites of their focal origin are indicated in Table 4.
Because of the limited number of the values obtained for
the distinct brain regions, no further statistical analysis
was performed.
Discussion
Performance monitoring represents a critical executive
function of the human brain. A major resource in this
research is to investigate the correlates of error processing
and conflict monitoring. Distinguishing between errors
and conflict is made difficult by the fact that conflict
may accompany errors and vice versa. Nevertheless, recent
evidence suggests that error-monitoring can be
viewed as separate from conflict-monitoring, and that
both functions may be subserved by distinct neural substrates
(Braver et al., 2001; Garavan et al., 2003; Kiehl
et al., 2000; Swick & Turken, 2002; Ullsperger & von
Cramon, 2001). The Ne/ERN phenomenon is, then, generally
thought to reflect the activity of the neural system
responsible for error monitoring (Falkenstein et al.,
2000, 2004; Gehring et al., 1993; Masaki and Segalowitz,
2004). In recent years, an increased interest in the
neural sources of the Ne/ERN can be observed in the
effort to identify the neuroanatomical bases of an errordetection
system. Dipole modeling ERP analyses, congruent
with neuroimaging studies, have repeatedly provided
evidence that a crucial source of the Ne/ERN lies
in the FMW. It was recently suggested that the ACC is
responsible for the modulation of autonomic nervous
system activity, which has been shown to be recruited
during error processing (Hajcak, McDonald, & Simons,
2003). In agreement with all these findings, the most
consistent proof of the Ne/ERN generators in our study
was obtained from FMW structures. Ne/ERN sources
were repeatedly revealed in the caudal ACC (BA 32) and
its vicinity – within the medial frontal gyrus on the ventral
bank of paracingulate sulcus (BA 9). It is noteworthy
that clear-cut generators of error-related potentials were
revealed in the identical sub-area of the ACC in our previous
intracerebral study on performance monitoring
(Brázdil et al., 2002). Two of the indubitable Ne/ERN
sources were found in the present study on the ventral
bank of the cingulate sulcus, where the rostral cingulate
motor area (CMA) is located (see Figures 3, 5, and 6).
These findings are consistent with the suggestion by Holroyd
(2001) that Ne/ERN is elicited by the activity of
neurons located within the CMA, and also with the results
of a recent fMRI study in which the authors observed
preferential activation of the human homolog of
the CMAduring error processing (Ullsperger & von Cramon,
2001). Another region that clearly elicited Ne/ERN
in our subjects was the rostral ACC – BAs 24 and 32, and
the adjacent medial frontal gyrus (BA 10). Significant
activation in the rostral ACC during the commission of
the errors, compared with during correct trials, was congruently
demonstrated in two fMRI studies of a
Go/NoGo task (Kiehl et al., 2000; Menon et al., 2001).
The responsiveness of the rostral ACC to errors may reflect
the role that this area plays in emotional processes.
This hypothesis can simply explain a discrepancy between
our results and those of Ullsperger’s fMRI study.
In the flankers task used in Ullsperger’s fMRI experiment,
errors can be corrected by a second key press. This
may lead to a less negative affective valence of errors for
the participants (Ullsperger & von Cramon, 2001). The
different functional significance of the rostral ACC involvement
in error processing also supports our finding
of an unequivocal time delay of Ne/ERNs within the rostral
ACC compared to the more caudal sub-areas of ACC.
It makes sense that emotional processing of the error
would be slightly delayed after the specific error-detection.
Our results match well with the contemporary view
of the ACC function, placing this region in a unique position
to translate intentions into actions (Paus, 2001).
Certainly the ACC is not dedicated exclusively to one
process, but it is activated in a huge number of conditions.
Most are related neither to errors nor to response
conflict. The anterior cingulate has been proposed to be
an essential part of the “anterior attention system,” and
as such it is importantly involved in target detection
(Baudena et al., 1995; Brázdil et al., 1999; Halgren, Marinkovic
K, & Chauvel, 1998; Posner et al., 1988; Posner
& Peterson, 1990). At the same time, the rostral division
M. Brázdil et al.: Neural Sources of Error-Related Negativity 251
Hogrefe & Huber Publishers Journal of Psychophysiology 2005; Vol. 19(4):244–255
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
of the ACC has been linked to emotion, and the dorsal
region has been associated with cognition and higher-order
motor control (Bush, Luu, & Posner, 2000; Devinsky,
Morrel, & Vogt, 1995; Paus et al., 1993). Therefore,
it is very unlikely that just single a neural source of the
Ne/ERN will be present within the ACC. Instead, functionally
different sub-areas of the ACC (subserving different
cognitive/affective functions) have to be involved
in error processing and, hence, participate in the genesis
of scalp Ne/ERN. In addition to a specific error-detection
system and the neural substrate for emotional processing
of the error, other systems should be activated by error
trials within a performed Go/NoGo task (e.g., the attentional
system, the systems for response selection, movement
execution, etc.). On the other hand, most of these
systems are also activated by correct trials and, consequently,
they can be reflected in the scalp CRN.
Neural sources of the Ne/ERN were repeatedly observed
in our study within the pre-SMA. The significant
hemodynamic response in the pre-SMA also has been
described in most fMRI studies on performance monitoring,
but this response was not specific to errors. Rather,
it was sensitive to conflict processing and to response
inhibition (Braver et al., 2001; Garavan et al., 2003; Menon
et al., 2001; Ullsperger & von Cramon, 2001; van
Veen et al., 2001). There is, nevertheless, clear-cut recent
evidence that pre-SMA shows both error and conflict
effects in a Go/NoGo task, and that this FMW structure
can be significantly activated by incorrect responses
(Garavan et al., 2004). In this sense, the pre-SMA very
likely contributes to the creation of the Ne/ERN phenomenon.
This statement is entirely congruent with our re-
sults.
Another brain region where undisputed Ne/ERN neural
sources were revealed in this study is the lateral prefrontal
cortex (PFC), namely the middle and inferior
frontal gyri. In the relevant literature, the congruently
lateral PFC has been shown to be involved in error processing.
The importance of this region and of the lateral
PFC-ACC interactions for the genesis of the Ne/ERN
was established in a study published by Gehring and
Knight (2000). Several neuroimaging studies also found
lateral prefrontal activations during errors (Carter et al.,
1998; Kiehl et al., 2000; Menon et al., 2001). Obvious
generators of error-related potentials were also detected
in the lateral PFC in our previous intracerebral study
(Brázdil et al., 2002). Because of the widespread reciprocal
cortico-cortical connections of the lateral PFC with
the supracallosal cingulate areas 24 and 32, these findings
are not truly surprising. Actually, coactivations of
the lateral PFC and the ACC during the performance of
a variety of tasks have been noticed frequently. Even if
the functional significance of such interactions remains
unclear, it was suggested that the lateral PFC computes
and maintains online information necessary for the selection
of an appropriate response, whereas the ACC facilitates
the implementation of the selected action (Paus
et al., 1993). Our results produced no evidence to support
the possible lateralization of the PFC involvement in error
processing.
Our observations of Ne/ERN sources in the orbitofrontal
cortex, the supramarginal gyrus, and the lateral
temporal neocortex are less consistent than findings
within the FMW and PFC structures. The orbito-frontal
cortex is known to be importantly involved in the modulation
of impulsivity (Bechara, Damasio, & Damasio,
2000; Fuster, 1995). This fits well with our evidence of
local Ne/ERN sources within BA 11 in two subjects.
However, neuroimaging studies have not yet revealed
any activation in this structure related to error processing.
On the other hand, the orbito-frontal cortex is a part
of the brain that is well known to be extremely prone to
susceptibility artifacts in fMRI experiments. Thus, false
negative findings in hemodynamic studies might be obtained.
In contrast, our separate proof of the neural
source of the Ne/ERN in the supramarginal gyrus (SMG)
is in agreement with the results of the Ullsperger and von
Cramon’s fMRI study (2001). But significant hemodynamic
response within a parietal lobe was described extremely
rarely in previous fMRI studies focused on
Go/NoGo tasks (Garavan et al., 2004). An important
contribution to the understanding of the neurophysiological
aspects of error processing may arise from our finding
of the shortest Ne/ERN latency within SMG. By
analogy, intracerebral investigations of P3 potential disclosed
the earliest involvement in the detection of targets
as that of the parietal lobe structures (Halgren et al.,
1998). Finally, the clear-cut latency differences between
the P3 and error-related potentials across distinct brain
lobes observed in our previous study suggest that the
Ne/ERN is also primarily elicited in parietal regions, and
only later in frontal and temporal regions (Brázdil et al.,
2002). The substantially different peak latencies in distinct
Ne/ERN neural sources seen in the present study
support that hypothesis and link different cognitive processes
in the human brain.
The role of the hippocampus in the processing of error
trials and the functional significance of the common local
activation after both erroneous hits and correct rejections
of rare stimuli should be discussed. The hippocampus
is known to be a structure importantly engaged in the
cognitive processing of external as well as internal stimuli.
It has also been identified as a powerful neural source
of some cognitive potentials including P3 (see Halgren
et al., 1998). This mesiotemporal structure is evidently
simultaneously involved in affective processing. In
terms of discussed emotional aspects in error processing,
the mesiotemporal structures (including the hippocam-
252 M. Brázdil et al.: Neural Sources of Error-Related Negativity
Journal of Psychophysiology 2005; Vol. 19(4):244–255 Hogrefe & Huber Publishers
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
pus) were coherently identified as additional generators
of error-related potentials in our previous study using an
oddball task. In that oddball study, erroneous motor responses
to frequent stimuli were analyzed and compared
to the correct motor responses to rare stimuli. A P3 potential
was observed after correctly hit targets, while errors
evoked undisputed Ne/Pe complexes in several
brain sites, with the most consistent proof of their generators
within the mesiotemporal region (Brázdil et al.,
2002). The design of the Go/NoGo task used in the present
study unfortunately impedes a meaningful assessment
of the role of the hippocampus in the genesis of
Ne/ERN. It could be expected that both error and correct
reactions to the rarely presented “K” stimuli would activate
hippocampal neurons and, thus, produce an ERP.
This was the case for all the hippocampal recordings.
However, an interpretation of the obtained ERP remains
speculative. Certainly neuronal populations responsible
for target detection would be activated in both conditions,
and, thus, the P3 potential should be present in both
STAs. On the other hand, neuronal populations engaged
in successful response inhibition and conflict monitoring
(and hence producing NoGo N2; Nieuwenhuis et al.,
2002) will also be involved in processing of the correctly
rejected “K”s and the distinct neuronal substrate for error
processing will produce similar Ne/ERNs after error trials.
If there is a brain region with a common involvement
of its neuronal populations both in error-processing and
response inhibition, then similar responses will be seen
in STAs. Even if it is conceivable that response inhibition
and error processing activate different brain areas, some
evidence has been accumulated from hemodynamic
studies that indicate that the error-processing network
indeed overlaps partly with the distributed network involved
in response inhibition. Brain regions that are involved
in both functions were identified to be the preSMA,
the caudal ACC, and the lateral prefrontal cortex
(Carter et al., 1998; Fiehler et al., 2004; Ford, Whitfield,
& Mathalon, 2004; Menon et al., 2001; Ullsperger & von
Cramon, 2001). Absolutely no information is available
for mesiotemporal regions that addresses their potential
role in conflict monitoring. The reason for this may be
the low sensitivity of fMRI methodology for the detection
of transient neural activity within the mesiotemporal
regions during some cognitive tasks. In contrast, this
“hidden” activity can be clearly detected in intracerebral
ERP recordings (Brázdil et al., 2005). Given the knowledge
currently available, one can assume that analogous
intracerebral ERPs in stimulus-triggered averages within
the hippocampi reflect the activation of the hippocampal
neural substrate for target detection, as well as the involvement
of hippocampal neurons in error processing,
and possibly response inhibition also. The hippocampal
P3 response is extremely strong and, thus, the Ne/ERN,
and possibly the NoGo N2, are still covered by this P3
response. An analogous situation can be seen in a few
other generators of Ne/ERN-like potentials, with identical
responses to error hits and correct rejections in STAs
(i.e., ACC /1/, medial frontal gyrus /1/, orbito-frontalcortex
/1/, and middle temporal gyrus /2/). An intracerebral
ERP study using a Go/NoGo task with equal rates of Go
and NoGo stimuli is suggested to eliminate the problem
of this P3 effect on Ne/ERN.
To summarize, our results clearly confirmed the key
role of ACC and the adjacent FMW in the genesis of
error-related negativity. However, our results show that
error processing involves an extensively distributed network
of different brain regions. In addition to FMW, other
cortical structures were proven to be engaged – the
lateral and basal prefrontal cortex, the lateral temporal
neocortex, and the supramarginal gyrus. At the same
time, it can be expected that the brain’s error-checking
systems include some other regions that are involved in
parallel fashion both in error processing and conflict
monitoring, but still participating in the genesis of scalp
Ne/ERN.
Acknowledgments
The study was supported by the MSMT Program. Also,
we wish to express our thanks to the neurosurgeons Z.
Novák and J. Chrastina for their collaboration.
References
Baudena, P., Halgren, E., Heit, G., & Clarke, J.M. (1995). Intracerebral
potentials to rare target and distractor auditory and
visual stimuli. III. Frontal cortex. Electroencephalography and
Clinical Neurophysiology, 94, 251–264.
Bechara, A., Damasio, H., Damasio, A.R. (2000). Emotion, decision
making and the orbitofrontal cortex. Cerebral Cortex, 10,
295–307.
Botvinick, M.M., Braver, T.S., Carter, C.S., Barch, D.M., & Cohen,
J.D. (2001). Evaluating the demand for control: Anterior
cingulate cortex and crosstalk monitoring. Psychological Review,
108, 624–652.
Braver, T.S., Barch, D.M., Gray, J.R., Molfese, D.L., & Snyder, A.
(2001). Anterior cingulate cortex and response conflict: Effects
of frequency, inhibition, and errors. Cerebral Cortex, 11,
825–836.
Brázdil, M., Dobšík, M., Mikl, M., Hluštík, P., Daniel, P., Pazourková,
M., Krupa, P., & Rektor, I. (2005). Combined event-related
fMRI and intracerebral ERP study of an auditory oddball
task. NeuroImage, 26(1), 285–293.
Brázdil, M., Rektor, I., Dufek, M., Daniel, P., Jurák, P., & Kuba,
R. (1999). The role of frontal and temporal lobes in visual discrimination
task – Depth ERP studies. Neurophysiologie
Clinique, 29, 339–350.
Brázdil, M., Roman, R., Falkenstein, M., Daniel, P., Jurák, P., &
Rektor, I. (2002). Error processing – evidence from intracereM.
Brázdil et al.: Neural Sources of Error-Related Negativity 253
Hogrefe & Huber Publishers Journal of Psychophysiology 2005; Vol. 19(4):244–255
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
bral ERP recordings. Experimental Brain Research, 146,
460–466.
Bush, G., Luu, P., & Posner, M.I. (2000). Cognitive and emotional
influences in anterior cingulate cortex. Trends in Cognitive Science,
4, 215–222.
Carter, C.S., Braver, T.S., Barch, D.M., Botvinick, M.M., Noll, D.,
& Cohen, J.D. (1998). Anterior cingulate cortex, error detection,
and the online monitoring of the performance. Science,
280, 747–749.
Coles, M.G.H., Scheffers, M.K., & Holroyd, C. (2000). Why is
there an ERN or Ne on correct trials? Psychophysiology, 37,
S9.
Dehaene, S., Posner, M.I., & Tucker, D.M. (1994). Localization
of a neural system for error detection and compensation. Psychological
Science, 5, 303–305.
Devinsky, O., Morrel, M.J., & Vogt, B.A. (1995). Contributions
of anterior cingulate cortex to behavior. Brain, 118, 279–306.
Falkenstein, M., Hohnsbein, J., & Hoormann, J. (1991). Effects of
crossmodal divided attention on late ERPcomponents. II. Error
processing in choice reaction tasks. Electroencephalography
and Clinical Neurophysiology, 78, 447–455.
Falkenstein, M., Hoormann, J., Christ, S., & Hohnsbein, J. (2000).
ERP components on reaction errors and their functional significance:
A tutorial. Biological Psychology, 51, 87–107.
Falkenstein, M. (2004). ERPcorrelates of erroneous performance.
In M. Ullsperger & M. Falkenstein (Eds.), Errors, conflicts,
and the brain. Current opinions on performance monitoring
(pp. 5–14). MPI special issue in human cognitive and brain
sciences 1. Leipzig: Max-Planck-Institute für Kognitions- und
Neurowissenschaften.
Fiehler, K., Ullsperger, M., & von Cramon, D.Y. (2004). Neural
correlates of error detection and error correction: Is there a
common neuroanatomical substrate? European Journal of
Neuroscience, 19, 3081–3087.
Ford, J.M. (1999). Schizophrenia: The broken P300 and beyond.
Psychophysiology, 36, 667–682.
Ford, J.M., Whitfield, S.L., & Mathalon, D.H. (2004). The neuroanatomy
of conflict and error: ERPand fMRI. In M. Ullsperger
& M. Falkenstein (Eds.), Errors, conflicts, and the brain. Current
opinions on performance monitoring (pp. 42–48). MPI
special issue in human cognitive and brain sciences 1. Leipzig:
Max-Planck-Institute für Kognitions- und Neurowissenschaf-
ten.
Fuster, J.M. (1995). Memory and planning. Two temporal perspectives
of frontal lobe function. In H.H. Jasper, S. Riggio, &
P.S. Goldman-Rakic (Eds.), Epilepsy and the functional anatomy
of the frontal lobe (pp. 9–20). New York: Raven.
Garavan, H., Ross, T.J., Murphy, G., Roche, R.A., & Stein, E.A.
(2002). Dissociable executive functions in the dynamic control
of behavior: Inhibition, error detection, and correction. NeuroImage,
17, 1820–1829.
Garavan, H., Ross, T.J., Kaufman, J., & Stein, E.A. (2003). A
midline dissociation between error-processing and responseconflict
monitoring. NeuroImage, 20, 1132–1139.
Garavan, H., Hester, R., & Fassbender, C. (2004). The impact of
individual differences and prefrontal control on action monitoring
revealed through fMRI. In M. Ullsperger & M. Falkenstein
(Eds.), Errors, conflicts, and the brain. Current opinions
on performance monitoring (pp. 48–55). MPI special issue in
human cognitive and brain sciences 1. Leipzig: Max-PlanckInstitute
für Kognitions- und Neurowissenschaften.
Gehring, W.J., Goss, B., Coles, M.G.H., Meyer, D.E., & Donchin,
E. (1993). A neural system for error detection and compensation.
Psychological Science, 4, 385–390.
Gehring, W.J., & Knight, R.T. (2000). Prefrontal-cingulate interactions
in action monitoring. Nature Neuroscience, 3,
516–520.
Hajcak, G., McDonald, N., & Simons, R.F. (2003). To err is autonomic:
Error-related brain potentials, ANS activity, and posterror
compensatory behavior. Psychophysiology, 40, 895–903.
Halgren, E., Marinkovic, K., & Chauvel, P. (1998). Generators of
the late cognitive potentials in auditory and visual oddball
tasks. Electroencephalography and Clinical Neurophysiology,
106, 156–164.
Holroyd, C.B., Dien, J., & Coles, M.G.H. (1998). Error-related
scalp potentials elicited by hand and foot movements: Evidence
for an output-independent error-processing system in
humans. Neuroscience Letters, 242, 65–68.
Holroyd, C.B. (2001). Reinforcement learning and the error-related
negativity: A computational and neurophysiological investigation.
Unpublished PhD dissertation, University of Illinois,
Urbana-Champaign.
Kiehl, K.A., Liddle, P.F., & Hopfinger, J.B. (2000). Error processing
and the rostral anterior cingulate: An event-related fMRI
study. Psychophysiology, 37, 216–223.
Luu, P., Flaisch, T., & Tucker, D.M. (2000). Medial frontal cortex
in action monitoring. The Journal of Neuroscience, 20,
464–469.
Masaki, H., & Segalowitz, S. (2004). Error negativity: A test of
the response conflict versus error detection hypotheses. In M.
Ullsperger & M. Falkenstein (Eds.), Errors, conflicts, and the
brain. Current opinions on performance monitoring (pp. 76–
83). MPI special issue in human cognitive and brain sciences
1. Leipzig: Max-Planck-Institute für Kognitions- und Neuro-
wissenschaften.
Menon, V., Adleman, N.E., White, C.D., Glover, G.H., & Reiss,
A.L. (2001). Error-related brain activation during a Go/NoGo
response inhibition task. Human Brain Mapping, 12, 131–143.
Miltner, W.H.R., Lemke, U., Holroyd, C., Scheffers, M.K., &
Coles, M.G.H. (1998). Where does the brain process errors?
On the neural generators of the ERN. Psychophysiology, 35, 7.
Nieuwenhuis, S., Ridderinkhof, K.R., Talsma, D., Coles, M.G.,
Holroyd, C.B., Kok, A., & van der Molen, M.W. (2002). A
computational account of altered error processing in older age:
Dopamine and the error related negativity. Cognitive, Affective
& Behavioral Neuroscience, 3, 17–26.
Paus, T., Petrides, M., Evans, A.C., & Meyer, E. (1993). Role of
the human anterior cingulate cortex in the control of oculomotor,
manual, and speech responses: Apositron emission tomography
study. Journal of Neurophysiology, 2, 453–469.
Paus, T. (2001). Primate anterior cingulate cortex: Where motor
control drive and cognition interface. Nature Reviews Neuroscience,
2, 417–424.
Posner, M.I., Petersen, S.E., Fox, P.T., & Raichle, M.E. (1988).
Localization of cognitive operations in the human brain. Science,
240, 1627–1631.
Posner, M.I., & Petersen, S.E. (1990). The attention system of the
human brain. Annual Review of Neuroscience, 13, 25–42.
Ridderinkhof, K.R., Ullsperger, M., Crone, E.A., & Nieuwenhuis,
S. (2004). The role of the medial frontal cortex in cognitive
control. Science, 306, 443–447.
Swick, D., & Turken, A.U. (2002). Dissociation between conflict
254 M. Brázdil et al.: Neural Sources of Error-Related Negativity
Journal of Psychophysiology 2005; Vol. 19(4):244–255 Hogrefe & Huber Publishers
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
detection and error monitoring in the human anterior cingulate
cortex. Proceedings of the National Academy of Sciences USA,
25, 16354–16359.
Talairach, J., Szikla, G., Tournoux, P., Prosalentis, A., Bordas-Ferrer,
M., & Covello, J. (1967). Atlas d’Anatomie Stereotaxique
du Telencephale. Paris: Masson.
Ullsperger, M., & von Cramon, D.Y. (2001). Subprocesses of performance
monitoring: A dissociation of error processing and
response competition revealed by event-related fMRI and
ERPs. NeuroImage, 14, 1387–401.
Ullsperger, M., & von Cramon, D.Y. (2004). Neuroimaging of
performance monitoring: Error detection and beyond. Cortex,
40, 593–604.
Van Veen, V., Cohen, J.D., Botvinick, M.M., Stenger, V.A., &
Carter, C.S. (2001). Anterior cingulate cortex, conflict monitoring,
and levels of processing. NeuroImage, 14, 1302–1308.
Van Veen, V., & Carter, C.S. (2002a). The anterior cingulate as a
conflict monitor: fMRI and ERP studies. Physiology & Behavior,
77, 477–82.
Van Veen, V., & Carter, C.S. (2002b). The timing of action-monitoring
processes in the anterior cingulate cortex. Journal of
Cognitive Neuroscience, 14, 593–602.
Vaughan, H.G., Weinberg, H., Lehmann, D., & Okada, Y. (1986).
Approaches to defining the intracranial generators of event-related
electrical and magnetic fields. In W.C. McCallum, R.
Zappoli, &F. Denoth (Eds.), Cerebral psychophysiology: Studies
in event-related potentials (EEG Suppl. 38, pp. 505–544).
Amsterdam: Elsevier Science Publ.
Vidal, F., Hasbroucq, T., & Bonnet, M. (2000). Is the “error negativity”
specific to errors? Biological Psychology, 51, 109–
128.
Yeung, N., Cohen, J.D., & Botvinick, M.M. (2004). The neural
basis of error detection: Conflict monitoring and the error-related
negativity. Psychological Review, 111, 931–959.
Accepted for publication: June 29, 2005
Address for correspondence
Milan Brázdil
First Department of Neurology, Masaryk University
St. Anne’s Hospital, Pekarska 53
656 91 Brno
Czech Republic
Tel. +420 543 182-639
Fax +420 543 182-624
E-mail mbrazd@med.muni.cz
M. Brázdil et al.: Neural Sources of Error-Related Negativity 255
Hogrefe & Huber Publishers Journal of Psychophysiology 2005; Vol. 19(4):244–255
ThisdocumentiscopyrightedbytheAmericanPsychologicalAssociationoroneofitsalliedpublishers.
Thisarticleisintendedsolelyforthepersonaluseoftheindividualuserandisnottobedisseminatedbroadly.
Příloha č. 6
Damborska A, Roman R, Brazdil M, Rektor I, Kukleta M. Post-movement processing in
visual oddball task - Evidence from intracerebral recording. Clinical Neurophysiology
2016;127(2):1297-1306. IF (2016) = 3,866
Post-movement processing in visual oddball task – Evidence from
intracerebral recording
Alena Damborská a,b
, Robert Roman a,b,⇑
, Milan Brázdil a,c
, Ivan Rektor a,c
, Miloslav Kukleta a
a
CEITEC – Central European Institute of Technology, Masaryk University, Brno, Czech Republic
b
Department of Physiology, Faculty of Medicine, Masaryk University, Brno, Czech Republic
c
1st Department of Neurology, St. Anne’s Faculty Hospital, Masaryk University, Brno, Czech Republic
a r t i c l e i n f o
Article history:
Accepted 25 August 2015
Available online 4 September 2015
Keywords:
Intracerebral EEG
ERP
Movement
Monitoring
Correct performance
Error
h i g h l i g h t s
 Multiple cortical structures are activated after correct motor performance including mesiotemporal
structures, anterior midcingulate, prefrontal, and temporal cortices.
 Equivalent involvement of these structures in task-variant nonspecific and target specific processes is
suggested.
 Networks generating correct and incorrect performance-related potentials probably share some common
nodes.
a b s t r a c t
Objective: To identify intracerebral sites activated after correct motor response during cognitive task and
to assess associations of this activity with mental processes.
Methods: Intracerebral EEG was recorded from 205 sites of frontal, temporal and parietal lobes in 18
epileptic patients, who responded by button pressing together with mental counting to target stimuli
in visual oddball task.
Results: Post-movement event-related potentials (ERPs) with mean latency 295 ± 184 ms after movement
were found in all subjects in 64% of sites investigated. Generators were consistently observed in
mesiotemporal structures, anterior midcingulate, prefrontal, and temporal cortices. Task-variant nonspecific
and target specific post-movement ERPs were identified, displaying no significant differences
in distribution among generating structures. Both after correct and incorrect performances the postperformance
ERPs were observed in frontal and temporal cortices with latency sensitive to error commission
in several frontal regions.
Conclusion: Mesiotemporal structures and regions in anterior midcingulate, prefrontal and temporal cortices
seem to represent integral parts of network activated after correct motor response in visual oddball
task with mental counting. Our results imply equivalent involvement of these structures in task-variant
nonspecific and target specific processes, and suggest existence of common nodes for correct and incorrect
responses.
Significance: Our results contribute to better understanding of neural mechanisms underlying goaldirected
behavior.
Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights
reserved.
1. Introduction
It is widely accepted that successive components of eventrelated
potentials (ERPs) elicited during a cognitive task are related
to successive stages in processing. Therefore, electrophysiological
recording can be used to monitor the probable location, timing
and intensity of brain activation during the task (Halgren et al.,
1998). Recently, an increased interest in the neural sources of ERPs
elicited in the post-performance period has been observed with the
aim to identify anatomical structures engaged in performance
monitoring. A well-known event-related potential provoked
by errors, error-related negativity (Ne/ERN) typically peaking
http://dx.doi.org/10.1016/j.clinph.2015.08.014
1388-2457/Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
⇑ Corresponding author at: Department of Physiology, Faculty of Medicine,
Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic. Tel.: +420 549 496
818.
E-mail address: roman@med.muni.cz (R. Roman).
Clinical Neurophysiology 127 (2016) 1297–1306
Contents lists available at ScienceDirect
Clinical Neurophysiology
journal homepage: www.elsevier.com/locate/clinph
100–150 ms after an erroneous response (Falkenstein et al., 1990;
Gehring et al., 1993) was proved by intracerebral studies to stem
from activation of multiple sources, with the most consistent
involvement being that of frontomedian wall and mesiotemporal
lobe structures (Brázdil et al., 2002, 2005; Pourtois
et al., 2010). A positive deflection occurring 200–500 ms after an
incorrect response and following the Ne/ERN referred to as error
positivity (Pe; Falkenstein et al., 1991, 1995, 2000) was suggested
to be generated in the anterior cingulate cortex (ACC; Herrmann
et al., 2004) and to have a common origin with the Ne/ERN
(Brázdil et al., 2002). In patients and also in healthy subjects the
so-called ‘‘correct response-related negativity” (Nc/CRN) is often
seen as negative deflection following correct responses in surface
electrophysiological recordings (Bonnefond et al., 2011; Coles
et al., 2001; Ford, 1999; Gehring and Knight 2000; Scheffers and
Coles, 2000; Vidal et al., 2000, 2003). Surprisingly, few studies
carefully investigated neural sources of Nc/CRN. This ERP was
observed at frontal (Mathalon et al., 2002; Meckler et al., 2011)
and frontocentral electrodes (Falkenstein et al. 2000; Hajcak
et al., 2005), and, therefore, generators in the rostral cingulate zone
were suggested (Carter et al., 1998; Roger et al., 2010). Similarly to
erroneous responses a small positivity can also be found for correct
responses called the correct response positivity (Pc), but it has generally
been used only as a baseline comparison for the Pe and little
has been written about it (Bates et al., 2004; Mathalon et al., 2002).
We believe, however, that studies focused on correct response processing
could contribute to better understanding of neural mechanisms
underlying successful goal-directed behavior. In addition to
these theoretical implications, studying the functional significance
of post-performance ERPs might also help to understand deficits in
action control and behavior adaptation observed in psychopathological
and neurological conditions (Taylor et al., 2007).
Scalp-recorded event-related potentials (ERPs) elicited during
the so-called ‘‘oddball” task have been employed for decades as a
useful tool for studying cognition processes. In the oddball task
the subject responds by button pressing and/or mental counting
only to the infrequent ‘‘target” stimulus, which is presented randomly
and repeatedly among frequent ‘‘nontarget” stimuli. Two
types of correct performance are observed – motor response in
the target variant (correct hit) and refraining from movement in
the nontarget variant (correct rejection). The oddball paradigm
can also induce two types of errors – response omission in the target
variant (incorrect rejection) and erroneous motor response
(false alarm) in the nontarget variant. As such, the analysis of ERPs
recorded in the post-performance period of this simple cognitive
task seems to be very useful for studying mechanisms of performance
monitoring. Previous intracerebral studies employing
motor tasks, however, focused rather on P3-like potentials
(Rektor et al. 2007) or ERPs appearing before and during a simple
acral limb movement (Rektor et al. 1998; Rektor, 2000) and did
not investigate ERP components that might appear later. To our
knowledge, the activity following correct responses in motor tasks
has not been systematically examined by any intracerebral study
yet.
While multiple sources were proved for event-related activity
in the post-performance period provoked by errors (Brázdil et al.,
2002, 2005; Pourtois et al., 2010) we have hypothesized whether
also in correct reactions to target stimuli of visual oddball task
multiple brain regions might be involved in processes that take
place there. If so, ERP activity in the post-movement period of
correctly performed task might be observed within a large-scale
neuronal network rather than limited to a single brain structure.
To this end, the present intracerebral ERP study was designed to
identify the cerebral sites consistently activated after correct
motor reactions in a visual oddball task. Therefore we assessed
across examined brain structures the occurrence frequency of
post-movement ERP evoked during the target task variant. In an
attempt to associate these ERPs with underlying mental processes,
we decided to evaluate these potentials in relation to ERPs elicited
after correct performance of nontarget task variant and after both
types of errors induced during the task. If the post-movement ERP
reflects different mental processes then in sites where this ERP was
detected, differences in post-performance activity in the nontarget
task variant should be observed. If the brain network related to
incorrect responses shares some common nodes with that related
to correct responses, then sites active in both conditions should
be identified. To explore the character of mental processes underlying
the post-movement ERPs we evaluated their latency and distribution
among brain structures.
2. Methods and materials
2.1. Subjects
Eighteen patients (14 men) aged from 23 to 45 years (median
30) were employed in the study (Table 1). All subjects suffered
from medically intractable epilepsy and were candidates for surgical
treatment. They all were under antiepileptic drug therapy,
which was determined by clinical considerations. During the period
of diagnostic examination by intracerebral EEG recording, the
doses of medicaments were reduced to allow seizures to develop
spontaneously. All patients had normal or corrected-to-normal
vision. The subjects gave us their informed consent to the experimental
protocol that had been approved by the Ethical Committee
of Masaryk University.
2.2. Experimental task
A visual oddball task was performed. The patients were sitting
comfortably in a moderately lighted room and were focusing on
the center of a monitor situated at about 100 cm from their eyes.
Yellow capital letters X (target) or O (nontarget) appeared repeatedly
on white background in random order as experimental stimuli.
Each stimulus presentation lasted 200 ms and the
interstimulus interval varied randomly between 2 and 5 s. The target
stimuli were five times less frequent than the nontarget ones.
The subjects were instructed to press a microswitch button with
Table 1
Patient characteristics and lobes investigated.
Patient Sex*
Age Lobes investigated *
Number of sites
investigated
1 M 20 RFT, LFT 11
2 M 28 RFT, LT 7
3 M 37 RFT, LFT 13
4 M 45 RFT, LFT 15
5 M 30 RFT, LFT 14
6 F 31 RFT 12
7 M 32 RFT, LTF 4
8 M 30 RFT, LFT 10
9 M 19 RF, LF 12
10 F 31 RFT, LFT 14
11 F 27 RT 6
12 M 19 RT, LT 14
13 M 41 RF, LF 8
14 M 25 RFT, LFT 17
15 M 34 RFTP 14
16 M 23 RFT, LFT 16
17 F 28 LFT 10
18 M 27 LT 8
*
M = male; F = female; R = right, L = left; F = frontal, T = temporal, P = parietal.
1298 A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306
the dominant hand as quickly as possible, whenever a target stimulus
appeared, to mentally count the target stimuli, and to ignore
the nontarget stimuli.
2.3. Data acquisition
Electrical activity was recorded during the task simultaneously
from various brain sites by means of standard Micro Deep semiflexible
multicontact platinum electrodes. Having a diameter of
0.8 mm, each electrode carried 5–15 contacts 2.0 mm long separated
by constant intervals of 1.5 mm. Strictly for diagnostic reasons
113 intracerebral depth electrodes were implanted into the
structures of the frontal, temporal, and parietal lobes (Table 1).
Every patient received 2–10 such electrodes exploring either or
both hemispheres. Long electrodes examined both lateral and
mesial cortical regions. The electrodes were placed using the
methodology of Talairach et al. (1967) and their position was afterwards
verified by magnetic resonance imaging with electrodes
in situ. The registration was made with the help of a 64-channel
Brain Quick EEG system (Micromed). All the recordings were
monopolar with respect to a reference electrode attached to the
right processus mastoideus. The impedances used were less than
5 kO. The EEG signal was amplified with a bandwidth of
0.1–40 Hz at a sampling rate of 128 Hz.
2.4. Analysis
The EEG signal was analyzed offline with the help of ScopeWin
software. The recordings from lesions and epileptogenic zones and
the trials with artefacts were rejected offline with visual inspection
made by two experienced persons. Switching the button in
response to a nontarget stimulus or its omission in response to a
target stimulus was considered as error. In each subject all
artefact-free trials with correct and at least 10 incorrect performances
were used for calculation of average curves. Exclusion of
a different number of trials and commission of a different number
of errors explains the interindividual variability in the number of
trials used for each average curve (target variant: 41–80 trials with
correct hits, 14 and 25 trials with incorrect refraining from movement;
nontarget variant: 191–342 trials with correct refraining
from movement, and 13, 14 and 18 trials with false alarms).
Post-movement ERP waves in the latency range of 0–900 ms from
movement were analyzed. Periresponse EEG periods (from À900 to
+900 ms from the movement onset) were averaged for target correct
responses using movement onset as a trigger. Peristimulus
EEG periods (from À300 to +1500 ms from the stimulus onset)
were averaged separately for target and nontarget correct and
incorrect performances using the stimulus onset as a trigger. The
statistical significance of ERP waves was computed between the
mean amplitude observed during the baseline period (from À600
to À100 ms from the stimulus onset) and the mean value computed
as a mean from the neighborhood of each point (170 ms
length) after stimuli or responses using a nonparametric Wilcoxon
Rank Sum (Signed Rank) test for paired samples. Records from one
contact of each multicontact intracerebral electrode implanted in a
particular anatomical structure were included in the analysis
selecting the one with the largest amplitude of ERP. The intracerebral
findings of steep voltage gradients uniquely proved the focal
origin of the waveform (Fig. 1).
The statistical analysis of recorded potentials was performed by
comparing the presence with the absence of the post-movement
ERP in correct reactions to target stimuli. The frequency differences
between brain structures were examined using the binomial test.
Involvement of these structures after correct performance in nontarget
task variant was compared using Fischer’s exact test. To test
the statistical significance of differences in latency of ERPs the
t-test for independent samples was used. In all tests the threshold
for statistical significance was set to P < 0.05.
3. Results
The performance of the subjects during the task was very accurate;
only three subjects (Nos. 5, 9, and 12) committed higher number
of errors (Table 2). The mean reaction time in correct responses
ranged from 457 ± 34 ms to 644 ± 78 ms (median 522 ms). In the
response-triggered averages (RTAs) of correct reactions to target
stimulus a prominent event-related potential in the postperformance
period (see Fig. 2) was detected in all subjects in
131 sites, i.e. 64% of sites investigated, with no significant difference
in localization within well examined frontal and temporal
lobes. These sites were distributed among multiple brain structures
(see Table 3); involving mesiotemporal structures, lateral
temporal, prefrontal, cingulate, frontal and parietal cortices, and
basal ganglia. The occurrence of the post-movement ERP in enough
examined (at least 5 investigated sites) brain structures ranged
from 43% to 100% (mean 67 ± 15%) of investigated sites. The binomial
test revealed no significant differences in frequency of findings
of post-movement ERP among these eleven structures. The
only exception were basal ganglia where finding of postmovement
ERP was significantly more frequent than in several,
mostly temporal, sufficiently examined brain structures (Table 4),
and anterior midcingulate cortex with borderline significantly
higher post-movement ERP occurrence than in inferior temporal
gyrus (P = 0.0485). When we restrict our results to observation of
generators, however, the list of presumably consistently involved
structures is shorter (see Table 5). The data in Fig. 1 demonstrate
three regions exhibiting signs of a local generator. Most frequently
(53% of sites investigated) the post-movement ERP was generated
in amygdala. Quite frequently, with occurrence not significantly
different from amygdala, generators were also observed in
Fig. 1. The response-triggered averaged ERPs for correct target trials recorded from
neighboring electrode contacts exhibiting steep voltage gradients in post-movement
period in right hippocampus in patient No. 2 (Z3–Z7), right anterior
midcingulate cortex in patient No. 9 (F1–F4), and right amygdala in patient No. 4
(A1–A3). Note the voltage variations across contacts with maximal peak amplitude
marked with black point (M = movement).
A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306 1299
hippocampus, parahippocampal and middle temporal gyri, anterior
midcingulate cortex, medial frontal, orbitofrontal, and dorsolateral
prefrontal gyri. On the other hand, despite sufficient
examination, the post-movement ERP was generated significantly
less frequently than in amygdala in superior and inferior temporal
gyri and no generator was found in basal ganglia. The comparison
between brain lobes revealed no significant differences in postmovement
ERP generator occurrence. In post-movement ERPs generated
in temporal and frontal lobes the mean peak latency relative
to motor response was 288.9 ± 171.6 ms and 248.7 ± 176.4 ms,
respectively. This difference, however, did not reach the statistical
significance.
The observed post-movement ERPs were mostly (80% sites)
monophasic, less frequently complex biphasic (14% sites) with
waveforms of opposite polarities, only occasionally (6% sites) two
separate waves of the same polarity were detected. In total, 157
ERP waves were observed in the post-movement period in the target
task variant. Their peak latency ranged from 14 to 726 ms
(mean 295 ± 184 ms, median 258 ms) with approximately half of
them (48%) appearing between 100 and 300 ms after movement
onset. Less than one third of post-movement ERPs (27%) were
observed in sites where no ERP component was detected during
the stimulus–response interval. This was revealed by the analysis
of stimulus-triggered averages (STAs) of correct reactions to target
stimulus. In these cases the post-movement ERP was observed as a
very late isolated ERP with a latency exceeding the mean reaction
time (e.g. in contact X0
8 in patient 3, see Fig. 3A); in the majority of
brain sites, however, this potential appeared only after components
elicited within the stimulus–response interval (e.g. in contact
D0
2 in patient 14, see Fig. 3A).
Approximately in 50% of the sites where post-movement ERP
was detected in the RTA of correct reactions to target stimuli, no
very late ERP was observed in the STA of correctly ignored nontarget
stimuli (e.g. examples in Fig. 3A). In the other half of the sites,
however, one (61 sites) or two (6 sites) potential waves with a
latency exceeding the subject’s mean reaction time were observed
(e.g. examples in Fig. 3B–D). The latency of all these 73 nontarget
post-performance ERP waves ranged from 500 to 1162 ms (mean
757 ± 168 ms) after the stimulus onset. A comparison of the first
nontarget with the first target post-performance ERP wave in the
STAs revealed that the latency of the nontarget ERP was shorter
by 174 ± 118 ms in 34 sites (Fig. 3B) and longer by 166 ± 88 ms
in 26 sites (Fig. 3C). In the remaining 7 sites the latencies were
almost identical differing by less than 20 ms (Fig. 3D).
In most anatomical structures examined both task-variant nonspecific
and target specific post-movement ERPs were observed
(Table 3). In the former case an ERP was detected both in target
post-movement and nontarget post-performance periods while in
the latter case the potential was found only in target task variant.
Mean latency of task-variant nonspecific (244.6 ± 142.1 ms) and
target specific (251.3 ± 164.5 ms) post-movement ERPs differed
only slightly and the differences did not reach a statistical significance.
Both the task-variant nonspecific and target specific postmovement
ERPs were observed in all but two patients. Only the
former or latter types of ERP were observed in subject No. 9 and
11, respectively. The occurrence frequency of task-variant nonspecific
and target specific ERP types did not significantly differ
between frontal and temporal lobes neither among individual
brain structures investigated. The only exception were the basal
Table 2
Number of artefact-free trials.
Patient Target Nontarget
Correct hit Incorrect rejection Correct rejection False alarm
1 49 0 281 3
2 45 0 236 1
3 62 6 266 1
4 41 5 221 1
5 65 1 332 18
6 57 3 328 2
7 49 0 261 0
8 52 0 247 0
9 77 25 343 14
10 46 3 267 0
11 59 1 243 2
12 80 14 282 13
13 53 3 201 0
14 61 1 316 4
15 63 1 242 0
16 66 3 305 3
17 57 0 222 0
18 57 2 270 1
Correct hit = motor response to target stimulus; Incorrect rejection = movement
omission after target stimulus; Correct rejection = refraining from movement in
nontarget task variant; False alarm = erroneous motor response in nontarget task
variant.
Fig. 2. Response-triggered averaged post-movement ERPs for correct target trials
(M = movement). Contact (subject), anatomical structure: T3 (15), right superior
temporal gyrus; C1 (3), right fusiform gyrus; A9 (3), right middle temporal gyrus;
C0
2 (4), left parahippocampal gyrus; G1 (13), right anterior cingulate cortex; A13
(6), right middle temporal gyrus; B0
7 (12), left superior temporal gyrus.
1300 A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306
ganglia in which only target specific ERPs were found (e.g. Fig. 3A)
and thus this result was significantly different from several brain
structures (see the last column in Table 4). No significant differences
in ERP type distribution, however, were found among generating
brain structures (see Table 5).
Errors, i.e. erroneously omitted (incorrect rejection) or erroneously
performed (false alarm) motor responses, were committed
quite frequently in three of the patients (Nos. 5, 9, and 12; see
Table 2). The subject’s mean reaction time was longer for false
alarms (RTf) compared to correct hits (RTc) by 56 ms, 23 ms, and
138 ms in patients Nos. 5, 9, and 12, respectively. In patient No.
5 no ERP was found after false alarms in sites where postmovement
ERP was recorded in correct reactions to target stimuli.
In both remaining patients, however, a very late ERP with latency
exceeding both RTc and RTf was observed in stimulus-triggered
Table 3
Distribution of ERPs across brain regions in correct reactions to target stimulus.
Anatomical structure Post-movement
ERPa
Sites examined/subjects Target specific
ERP/subjectsb
Task-variant nonspecific
ERP/subjectsc
Anterior midcingulate cortex 9(82) 11/7 3/3 6/5
Pregenual anterior cingulate cortex 2(67) 3/2 2/1 0/0
Rostro- and dorsomedial prefrontal cortices 6(75) 8/5 2/2 4/2
Orbitofrontal cortex 8(57) 14/8 6/5 2/2
Dorsolateral prefrontal cortex 15(71) 21/10 8/5 7/3
Supplementary motor area 0(0) 2/1 0/0 0/0
Premotor cortex 1(100) 1/1 1/1 0/0
Primary motor cortex 3(75) 4/2 0/0 3/2
Basal ganglia 7(100) 7/5 7/5 0/0
Amygdala 10(67) 15/12 5/5 5/5
Hippocampus 17(65) 26/15 10/8 7/6
Parahippocampal gyrus 6(55) 11/8 1/1 5/3
Fusiform gyrus 4(100) 4/5 1/1 3/3
Superior temporal gyrus 13(57) 23/11 5/5 8/4
Middle temporal gyrus 19(61) 31/15 9/6 10/7
Inferior temporal gyrus 6(43) 14/8 1/1 5/4
Lingual gyrus 1(50) 2/1 0/0 1/1
Posterior cingulate cortex 0(0) 2/1 0/0 0/0
Inferior parietal lobule 2(67) 3/1 1/1 1/1
Somatosensory cortex 2(67) 3/2 1/1 1/1
Frontal lobe 51(72) 71/14 29/13 22/7
Temporal lobe 76(60) 126/17 32/14 44/14
Parietal lobe 4(50) 8/2 2/1 2/1
Total 131(64) 205/18 63/17 68/17
a
Number of sites (% of sites examined) with positive observations of post-movement ERP in response-triggered averages of correct reactions to
target stimulus.
b
Number of sites/subjects with observed target post-movement ERP but with negative finding in post-performance period of nontarget task variant.
c
Number of sites where ERP was observed both in target post-movement and nontarget post-performance periods.
Table 4
Comparison between basal ganglia and other brain structures (P-values).
Anatomical structure Binominal test Fisher’s exact test
Anterior midcingulate cortex 0.2341 0.0114*
Rostro- and dorsomedial prefrontal cortices 0.1553 0.0210*
Orbitofrontal cortex 0.0400*
0.4667
Dorsolateral prefrontal cortex 0.1073 0.0513
Amygdala 0.0843 0.0441*
Hippocampus 0.0663 0.0648
Parahippocampal gyrus 0.0037*
0.0047*
Superior temporal gyrus 0.0341*
0.0147*
Middle temporal gyrus 0.0454*
0.0227*
Inferior temporal gyrus 0.0112*
0.0047*
*
P < 0.05, significant difference in numbers of post-movement ERPs (binominal
test) or in frequency of target specific and task-variant nonspecific ERPs (Fisher’s
exact test) between basal ganglia and other sufficiently examined brain structures.
Table 5
Distribution of generators across brain regions in correct reactions to target stimulus.
Anatomical structure Generatorsa
Target
specific
ERPb
Task-variant
nonspecific
ERPc
Anterior midcingulate cortex 5(42) 1 4
Pregenual anterior cingulate cortex 1(50) 1 0
Rostro- and dorsomedial prefrontal
cortices
2(25) 1 1
Orbito-frontal cortex 3(21) 2 1
Dorsolateral prefrontal cortex 8(38) 3 5
Supplementary motor area 0(0) 0 0
Premotor cortex 1(100) 1 0
Primary motor cortex 1(25) 0 1
Basal ganglia 0(0) 0 0
Amygdala 8(53) 4 4
Hippocampus 8(31) 4 4
Parahippocampal gyrus 5(45) 1 4
Fusiform gyrus 1(25) 1 0
Superior temporal gyrus 5(22) 3 2
Middle temporal gyrus 10(32) 5 5
Inferior temporal gyrus 2(14) 0 2
Lingual gyrus 0(0) 0 0
Posterior cingulate cortex 0(0) 0 0
Inferior parietal lobule 0(0) 0 0
Somatosensory cortex 1(33) 0 1
Frontal lobe 21 9 12
Temporal lobe 39 18 21
Parietal lobe 1 0 1
Total 61 27 34
Sufficiently examined brain structures (at least 5 sites investigated) with high
occurrence of generators are in bold format.
a
Number of sites (% of sites examined) displaying signs of proximity to structure
generating the target post-movement ERP either.
b
With negative finding in nontarget post-performance period,
c
With ERP observed in nontarget post-performance period.
A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306 1301
averages of erroneous responses. Tables 6 and 7 show distribution
of these ERPs across brain regions. Latencies of stimulus-triggered
averaged ERPs detected after correct and incorrect performances
are displayed in Table 8. In patient No. 9 the latency after incorrect
rejection or false alarm was either longer, with a delay of up to
469 ms (Fig. 4A), or almost identical, with a difference no greater
than 20 ms (Fig. 4B), compared to the latency observed after correct
hit or correct rejection, respectively. In patient No. 12 the
latencies after correct and incorrect performances differed less
than by 20 ms (Fig. 4B).
In an attempt to associate the post-movement ERP observed in
the target variant of a visual oddball task with underlying
mental processes the following competent characteristics of
post-performance ERPs were identified: (1) a waveform prominent
to target but absent to nontarget stimuli; (2) a waveform prominent
both to target and nontarget stimuli; (3) a longer latency of
ERP when motor response was erroneously omitted in reaction
to target stimulus; (4) a longer latency of ERP when motor
response was erroneously performed in reaction to nontarget
stimulus.
4. Discussion
This study examined the electrophysiological indicators of postperformance
brain activity. By using depth EEG recording during a
visual oddball task we have demonstrated that: (1) postmovement
ERPs in correct target trials were observed in multiple
cortical structures; (2) ERPs recorded after correct performance
were either specific for the target task variant or were observed
in both variants of the visual oddball task; (3) in several brain sites
both after correct and incorrect performances clear-cut ERP components
were observed; (4) in brain sites sensitive to error commission
the latency of the post-performance ERP was longer in
incorrect compared to correct performance.
4.1. Generators of post-performance ERPs in correctly performed
target trials are distributed in multiple brain regions
Two recent source localization studies on performance monitoring
focused on identifying ICA components reflecting the postperformance
ERPs for correct and incorrect responses (Hoffmann
Fig. 3. Post-performance ERP: (A) prominent in target but absent in nontarget trials; (B) with shorter latency in nontarget than in target trials; (C) with longer latency in
nontarget than in target trials; (D) with almost identical latency in target and nontarget trials. On the left side, response-triggered (M = movement) averaged ERPs for correct
target trials; on the right side, stimulus-triggered (S = stimulus) averaged ERPs for correct target (thick lines) and correct nontarget (thin lines) trials in the same contact. Note
that post-performance target ERP is either preceded by clear-cut earlier components or not, e.g. in D0
2(14) or X0
8(3) in section A, respectively. Contact (subject), anatomical
structure: D0
2(14), left superior temporal gyrus; X0
8(3), left putamen; M12(9), right primary motor cortex; O1(13), right orbitofrontal cortex; F8(16), right dorsolateral
prefrontal cortex; B8(3), right middle temporal gyrus; F2(9), right anterior cingulate cortex; A0
1(12), left superior temporal gyrus.
1302 A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306
and Falkenstein, 2010; Roger et al., 2010). Both studies, however,
aimed to identify ICA components related to the ERN/Ne, and the
component selection was based on error trials (Roger et al.,
2010) or on differences between error and correct trials
(Hoffmann and Falkenstein, 2010). To the best of our knowledge,
no source localization or even intracerebral study primarily aimed
at identifying sources of post-performance ERPs for correct
responses. Similarly, several EEG/fMRI studies identified correlations
between the hemodynamic response in the rostral cingulate
zone and scalp recorded ERN (for review see Ullsperger et al., 2014)
but none of them was focused on correct response-related activity.
From this point of view, the current study provides unique intracerebral
data suggesting the existence of multiple sources of ERPs
elicited after correct performance. We observed clear-cut correct
response-related activation in many different brain regions
(Table 3). Since converging observations seem to indicate that
the field created by neurons more than one centimeter away from
the recording site account for only a negligible portion of the
intracranially recorded EEG signal (Lachaux et al., 2003), these
ERPs might be considered to emerge from the brain structure in
which they were observed. Most convincingly, our results suggest
engagement of mesiotemporal structures, anterior midcingulate,
prefrontal and lateral temporal cortices, in which steep voltage
gradient across neighbouring contacts was observed as an
unequivocal evidence of a potential generation. Thus, besides
ACC whose involvement was previously documented by source
localization study (Roger et al., 2010) in correct response-related
negativity generation; our results suggest recruitment of several
other brain structures following correct motor responses. Our
results support the view, that the post-movement brain processes
in correct responses of visual oddball task are realized through
activation of large-scale neuronal networks. In the present study
most brain structures that form these networks were active in discrete
sites while in other sites no evoked electrophysiological
activity was recorded. No difference in occurrence frequency was
found among involved brain structures except for basal ganglia
where post-movement ERP was found in all examined sites. Due
to the fact that recording sites were selected according to diagnostic
concerns, and most regions were not explored, the possibility to
use this result for quantitative comparison of activation of different
brain structures is limited. Thus, it still remains in question as to
whether some structures generate the post-movement ERP in a
greater portion of its volume than other structures.
4.2. Parallel systems of correct performance processing
In all well examined structures except for basal ganglia, i.e. in
mesiotemporal structures, anterior midcingulate, prefrontal and
lateral temporal cortices, we observed task-variant nonspecific
post-movement ERP type. This suggests that the studied postmovement
ERP might reflect some processes that are supposed to
take place in both variants of the task, such as stimulus evaluation
Table 6
Distribution of ERPs across brain regions in correct reactions to target and erroneous
reactions to nontarget stimuli.
Anatomical structure Post-movement
ERPsa
Sites
examined/
subjects
False alarm/
subjectsb
Anterior cingulate cortex 3 4/2 0/0
Orbitofrontal cortex 0 2/1 –
Dorsolateral prefrontal cortex 3 3/1 3/1
Supplementary motor area 0 2/1 –
Premotor cortex 1 1/1 0/0
Primary motor cortex 2 3/1 2/1
Amygdala 1 1/1 0/0
Hippocampus 1 2/1 0/0
Parahippocampal gyrus 3 3/1 0/0
Superior temporal gyrus 7 9/2 1/1
Middle temporal gyrus 2 3/2 1/1
Inferior temporal gyrus 0 4/1 –
Lingual gyrus 1 2/1 0/0
Somatosensory cortex 0 1/1 –
Frontal lobe 9 15/2 5/1
Temporal lobe 15 24/2 2/1
Parietal lobe 0 1/1 –
Total 24 40/3 7/2
a
Number of sites with positive observations of post-movement ERP in responsetriggered
averages of correct reactions to target stimulus.
b
Number of sites/subjects with positive observation of ERP elicited both after
correct reactions to target and erroneous reactions to nontarget stimuli.
Table 7
Distribution of ERPs across brain regions in correct and erroneous reactions to target
stimuli.
Anatomical structure Post-movement ERPsa
Sites
examined/
subjects
Incorrect
rejection/
subjectsb
Anterior cingulate cortex 2 3/1 2/1
Dorsolateral prefrontal cortex 3 3/1 3/1
Supplementary motor area 0 2/1 –
Primary motor cortex 2 3/1 2/1
Hippocampus 1 2/1 0/0
Superior temporal gyrus 6 8/1 2/1
Middle temporal gyrus 2 2/1 1/1
Lingual gyrus 1 2/1 0/0
Somatosensory cortex 0 1/1 –
Frontal lobe 7 11/1 7/1
Temporal lobe 10 14/1 3/1
Parietal lobe 0 1/1 –
Total 17 26/2 10/2
a
Number of sites with positive observations of post-movement ERP in responsetriggered
averages of correct reactions to target stimulus.
b
Number of sites/subjects with positive observation of ERP elicited both after
correct and erroneous reactions to target stimuli.
Table 8
Latency (ms) of post-performance ERP measured from stimulus onset in two patients.
Anatomical
structure/contact
Target Nontarget
Correct
hit
Incorrect
rejection
Delay Correct
rejection
False
alarm
Delay
Patient No 9 (RTc = 474 ms, RTf = 497 ms)
aMCC/G1 617 1086 469 703 – –
aMCC/F2 601 1062 461 693 – –
DLPFC/G14 804 1062 258 742 875 133
DLPFC/F14 688 1070 382 742 968 226
DLPFC/F´9 601 1055 454 727 723 À4
PMC/M12 1046 1125 79 734 1117 383
PMC/M0
11 950 966 16 727 1009 282
Patient No 12 (RTc = 594 ms, RTf = 732 ms)
STG/T4 1018 999 À19 – – –
STG/T0
5 836 827 À9 774 789 15
MTG/D0
12 824 827 3 768 762 À6
Correct hit = motor response to target stimulus; Incorrect rejection = movement
omission after target stimulus; Correct rejection = refraining from movement in
nontarget task variant; False alarm = erroneous motor response in nontarget task
variant; Delay = the ERP latency difference between incorrect and correct performances;
RTc = mean reaction time for correct hit condition; RTf = mean reaction
time for false alarm condition; aMCC = anterior midcingulate cortex; DLPFC =
dorsolateral prefrontal cortex; PMC = primary motor cortex; STG = superior
temporal gyrus; MTG = middle temporal gyrus; 0
= left side; regions generating
post-movement ERP are written in italics format.
A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306 1303
processing, attentional processes or performance monitoring. This
interpretation is in line for instance with our previous study where
we provided evidence that hippocampal ERPs, either preceding or
following correct motor responses, could represent some stimulus
evaluation processing (Roman et al., 2013) or with widely accepted
role of frontal lobe for monitoring the effect of actions on the external
environment (Stuss and Benson, 1986; Stuss, 2011). Such a
monitoring process could just follow each correctly performed
task, irrespective of whether the required behavior was movement
execution and mental counting or movement inhibition and ignoring
of the stimuli.
The finding of target specific ERPs in all well examined brain
structures of frontal and temporal lobes may indicate their
involvement in counting related brain processes. Previous observations
of counting related brain activity in the lateral temporal neocortex
and ACC (Brázdil et al., 2003) support this view. Since higher
demands on memory functions in the target task variant with
mental counting are expected, participation of these structures in
memory processes might be suggested. Nevertheless, this type of
ERP could as well only reflect more attention or higher cognitive
load related to arithmetic without any specific relation to maintenance
of episodic information in short term memory. Some processes
underlying target specific activity may also be related to
previous movement execution. In case of basal ganglia, in which
we have found only target specific ERPs, this interpretation would
perfectly fit with their prominent role in motor functions. According
to another interpretation, this ERP could simply reflect passive
re-entrance of somatosensory signals or a post-performance resetting
of brain circuits without any elaboration or use of the performance
features.
Both the task variant nonspecific and target specific ERP types
were consistently generated in the same structures (see Table 5).
The mere existence of these two types of post-movement ERP
favors the view that after correct movement execution parallel
processing in at least two distinct systems takes place sharing at
least partially the same anatomical substrate. Parallel activation
of two different systems, i.e. performance-monitoring and
movement-monitoring, was also identified to be responsible for
Ne/ERN generation (Yordanova et al., 2004).
4.3. Common nodes of correct and incorrect behavior processing
In the present study we demonstrated that there are brain sites
in which both correct and incorrect performances may elicit a
prominent potential in the post-performance period. We found
sites in which ERPs were recorded both after correct motor
response and its erroneous omission in the target variant of the
task. In some of these sites we even recorded ERPs both after correct
refraining from movement and erroneous movement execution
in the nontarget task variant. Our findings suggest that
Fig. 4. Stimulus-triggered (S = stimulus) averaged ERPs for correct (thin line) and incorrect (thick line) trials. (A) Longer latency of post-performance ERP when motor
response was erroneously omitted in reaction to target stimulus (left section) or erroneously performed in reaction to nontarget stimulus (right section). (B) Almost identical
latency of post-performance ERP after correct and incorrect performance of target (left section) or nontarget (right section) task variant. Contact (subject), anatomical
structure: G1(9), right anterior cingulate cortex; F14(9), right dorsolateral prefrontal cortex; F2(9), right anterior cingulate cortex; F0
9(9), left dorsolateral prefrontal cortex;
T4(12), right superior temporal gyrus; D0
12(12), left middle temporal gyrus; M12(9), right primary motor cortex; G14(9), right dorsolateral prefrontal cortex; M0
11(9), left
primary motor cortex.
1304 A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306
cortical networks engaged in the generation of correct and incorrect
performance-related potentials could have some common
nodes. The fact that in several sites post-performance ERPs were
observed both in correct hit and false alarm conditions seems to
further support findings of recent studies using independent component
analysis, which strongly suggest that Ne/ERN and Nc/CRN
stem from largely overlapping generators (Roger et al., 2010;
Wessel et al., 2012). The demonstration of post-performance ERPs
in false alarms stands in line with a large body of literature on error
processing and supports the existence of an already known and
well-described Ne/ERN and Pe complex (Wessel, 2012). On the
other hand, our finding of post-performance ERPs after erroneous
movement omission is rather unique. Our finding suggests that
some kind of error processing might take place also after incorrect
movement inhibition. Actually, it should not be surprising that also
after this type of error some monitoring process might take place
that compares the predicted behavior with the actual one.
Another interesting finding of the present study is that there are
sites in anterior midcingulate, dorsolateral prefrontal and primary
motor cortices, where the error commission has an influence on
the latency of post-performance ERP with higher values for incorrect
performance (Fig. 4A). Erroneous refraining from movement in
the target variant of the task revealed a longer latency of the ERP
than the correct hit condition. Similarly, erroneously performed
movement in the nontarget variant of the task revealed a longer
latency than correct inhibition of motor response. If the observed
latency differences indicated that the preceding processes are prolonged
or higher in number in case of error commission, then both
correct and incorrect post-performance ERP components might
represent some final evaluation processes related to finished
action. Our findings of latency differences are analogous with those
reported in a recent intracerebral study (Pourtois et al., 2010). In a
go/nogo task the authors demonstrated local field potentials in the
amygdala with evident temporal unfolding. A monophasic ERP
around motor execution for correct hits was delayed by $300 ms
for false alarms, even though the actual reaction times were almost
identical in these two conditions. Our findings not only confirmed
typical involvement of ACC in action monitoring and cognitive control
(Ridderinkhof et al., 2004) but also suggested recruitment of
the dorsolateral prefrontal cortex and the primary motor cortex
in error detection mechanisms.
A striking result of our current study was activation of the primary
motor cortex in nonmotor task conditions, both after erroneous
and correct movement inhibition (Table 8 and Fig. 4). This
result extends the recent hypothesis of rostral premotorsubcortical
networks serving as a gateway between the cognitive
and motor networks (Hanakawa, 2011) suggesting that the primary
motor cortex might be also involved in processes not directly
associated with motor action.
5. Conclusions
Although the present study was limited by non-systematic
examination of the brain due to strictly diagnostic purposes of
electrode implantation, the results clearly suggest that besides
the anterior cingulate cortex, also the mesiotemporal structures,
and lateral temporal and prefrontal cortices are activated following
correct responses. The spatiotemporal characteristics, task-variant
specificity, and sensitivity to errors demonstrated that the
observed post-movement activity might code various pieces of
information needed for cognitive control, movement- and performance
monitoring, and evaluation processes. It appears, however,
that a one-to-one matching of post-movement ERP type and type
of information is beyond the limitations of our study.
Acknowledgments
Supported by the project ‘‘CEITEC – Central European Institute
of Technology” (CZ.1.05/1.1.00/02.0068) from the European Regional
Development Fund.
The funder had no role in study design, data collection and analysis,
decision to publish, or preparation of the manuscript.
Conflict of interest: None of the authors have potential conflicts
of interest to be disclosed.
References
Bates AT, Liddle PF, Kiehl KA, Ngan ETC. State dependent changes in error
monitoring in schizophrenia. J Psychiatr Res 2004;38:347–56.
Bonnefond A, Doignon-Camus N, Hoeft A, Dufour A. Impact of motivation on
cognitive control in the context of vigilance lowering: an ERP study. Brain Cogn
2011;77:464–71.
Brázdil M, Roman R, Falkenstein M, Daniel P, Jurák P, Rektor I. Error processing –
evidence from intracerebral ERP recordings. Exp Brain Res 2002;146:460–6.
Brázdil M, Roman R, Daniel P, Rektor I. Intracerebral somatosensory event-related
potentials: Effect of response type (button pressing versus mental counting) on
P3-like potentials within the human brain. Clin Neurophysiol
2003;114:1489–96.
Brázdil M, Roman R, Daniel P, Rektor I. Intracerebral error-related negativity in a
simple Go/NoGo task. J Psychophysiol 2005;19:244–55.
Carter CS, Braver TS, Barch DM, Botvinick MM, Noll D, Cohen JD. Anterior cingulate
cortex, error detection, and the online monitoring of performance. Science
1998;280:747–9.
Coles MGH, Scheffers MK, Holroyd CB. Why is there an ERN/Ne on correct trials?
Response representations, stimulus-related components, and the theory of
error-processing. Biol Psychol 2001;56:173–89.
Falkenstein M, Hohnsbein J, Hoormann J, Blanke L. Effects of errors in choice
reaction tasks on the ERP under focused and divided attention. In: Brunia CHM,
Gaillard AWK, Kok A, editors. Psychophysiological brain
research. Tilburg: Tilburg University Press; 1990. p. 192–5.
Falkenstein M, Hohnsbein J, Hoormann J, Blanke L. Effects of crossmodal divided
attention on late ERP components. II. Error processing in choice reaction tasks.
Electroencephalogr Clin Neurophysiol 1991;78:447–55.
Falkenstein M, Koshlykova NA, Kiroj VN, Hoormann J, Hohnsbein J. Late ERP
components in visual and auditory Go/Nogo tasks. Electroencephalogr Clin
Neurophysiol 1995;96:36–43.
Falkenstein M, Hoormann J, Christ S, Hohnsbein J. ERP components on reaction
errors and their functional significance: a tutorial. Biol Psychol
2000;51:87–107.
Ford JM. Schizophrenia: the broken P300 and beyond. Psychophysiology
1999;36:667–82.
Gehring WJ, Knight RT. Prefrontal-cingulate interactions in action monitoring. Nat
Neurosci 2000;3:516–20.
Gehring WJ, Goss B, Coles MGH, Meyer DE, Donchin E. A neural system for error
detection and compensation. Psychol Sci 1993;4:385–90.
Hajcak G, Moser JS, Yeung N, Simons RF. On the ERN and the significance of errors.
Psychophysiology 2005;42:151–60.
Halgren E, Marinkovic K, Chauvel P. Generators of the late cognitive potentials in
auditory and visual oddball tasks. Electroencephalogr Clin Neurophysiol
1998;106:156–64.
Hanakawa T. Rostral premotor cortex as a gateway between motor and cognitive
networks. Neurosci Res 2011;70:144–54.
Herrmann MJ, Römmler J, Ehlis AC, Heidrich A, Fallgatter AJ. Source localization
(LORETA) of the error-related-negativity (ERN/Ne) and positivity (Pe). Cogn
Brain Res 2004;20:294–9.
Hoffmann S, Falkenstein M. Independent component analysis of erroneous and
correct responses suggests online response control. Hum Brain Mapp
2010;31:1305–15.
Lachaux JP, Rudrauf D, Kahane P. Intracranial EEG and human brain mapping. J
Physiol Paris 2003;97:613–28.
Mathalon DH, Fedor M, Faustman WO, Gray M, Askari N, Ford JM. Responsemonitoring
dysfunction in schizophrenia: an event-related brain potential
study. J Abnorm Psychol 2002;111:22–41.
Meckler C, Allain S, Carbonnell L, Hasbroucq T, Burle B, Vidal F. Executive control
and response expectancy: a laplacian ERP study. Psychophysiology
2011;48:303–11.
Pourtois G, Vocat R, N’Diaye K, Spinelli L, Seeck M, Vuilleumier P. Errors recruit both
cognitive and emotional monitoring systems: simultaneous intracranial
recordings in the dorsal anterior cingulate gyrus and amygdala combined
with fMRI. Neuropsychologia 2010;48:1144–59.
Rektor I. Cortical activation in self-paced versus externally cued movements: a
hypothesis. Parkinsonism Relat D 2000;6:181–4.
Rektor I, Louvel J, Lamarche M. Intracerebral recording of potentials accompanying
simple limb movements: a SEEG study in epileptic patients. Electroencephalogr
Clin Neurophysiol 1998;107:277–86.
A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306 1305
Rektor I, Brázdil M, Nestrašil I, Bareš M, Daniel P. Modifications of cognitive and
motor tasks affect the occurrence of event-related potentials in the human
cortex. Eur J Neurosci 2007;26:1371–80.
Ridderinkhof KR, Ullsperger M, Crone EA, Nieuwenhuis S. The role of the medial
frontal cortex in cognitive control. Science 2004;306:443–7. http://dx.doi.org/
10.1126/science.1100301.
Roger C, Bénar CG, Vidal F, Hasbroucq T, Burle B. Rostral cingulate zone and correct
response monitoring: ICA and source localization evidences for the unicity of
correct- and error-negativities. NeuroImage 2010;51:391–403.
Roman R, Brázdil M, Chládek J, Rektor I, Jurák P, Sveˇtlák M, Damborská A, Shaw DJ,
Kukleta M. Hippocampal negative event-related potential recorded in humans
during a simple sensorimotor task occurs independently of motor execution.
Hippocampus 2013;23:1337–44.
Scheffers MK, Coles MGH. Performance monitoring in a confusing world: Errorrelated
brain activity, judgments of response accuracy, and types of errors. J Exp
Psychol Human 2000;26:141–51.
Stuss DT. Functions of the frontal lobes: relation to executive functions. J Int
Neuropsychol Soc 2011;17:759–65.
Stuss DT, Benson DF. Personality and emotions. The frontal lobes. New York: Raven
Press; 1986. pp 121–138.
Talairach J, Szikla G, Tournoux P, Prosalentis A, Bordas-Ferrer M, Covello J. Atlas
d’anatomie stéréotaxique du télencéphale. Paris: Masson; 1967.
Taylor SF, Stern ER, Gehring WJ. Neural systems for error monitoring: recent
findings and theoretical perspectives. Neuroscientist 2007;13:160–72.
Ullsperger M, Danielmeier C, Jocham G. Neurophysiology of performance
monitoring and adaptive behavior. Physiol Rev 2014;94:35–79.
Vidal F, Hasbroucq T, Grapperon J, Bonnet M. Is the ‘error negativity’ specific to
errors? Biol Psychol 2000;51:109–28.
Vidal F, Burle B, Bonnet M, Grapperon J, Hasbroucq T. Error negativity on correct
trials: a reexamination of available data. Biol Psychol 2003;64:265–82.
Wessel JR. Error awareness and the error-related negativity: evaluating the first
decade of evidence. Front Hum Neurosci 2012;6:88.
Wessel JR, Danielmeier C, Morton JB, Ullsperger M. Surprise and error: common
neuronal architecture for the processing of errors and novelty. J Neurosci
2012;32:7528–37.
Yordanova J, Falkenstein M, Hohnsbein J, Kolev V. Parallel systems of error
processing in the brain. NeuroImage 2004;22:590–602.
1306 A. Damborská et al. / Clinical Neurophysiology 127 (2016) 1297–1306
Příloha č. 7
Kukleta M, Bob P, Brazdil M, Roman R, Rektor I. The level of frontal-temporal beta-2 band
EEG synchronization distinguishes anterior cingulate cortex from other frontal regions.
Consciousness and Cognition 2010;19:879-886. IF (2010) = 2,027
The level of frontal-temporal beta-2 band EEG synchronization
distinguishes anterior cingulate cortex from other frontal regions
M. Kukleta a,c,*, P. Bob a,c
, M. Brázdil b
, R. Roman a
, I. Rektor b
a
Department of Physiology, Faculty of Medicine, Masaryk University, Brno, Czech Republic
b
First Department of Neurology, Faculty of Medicine, Masaryk University, Brno, Czech Republic
c
Department of Psychiatry, First Faculty of Medicine, Charles University, Prague, Czech Republic
a r t i c l e i n f o
Article history:
Received 2 November 2009
Available online 14 May 2010
Keywords:
Anterior cingulate cortex
Consciousness
EEG
Large-scale integration
a b s t r a c t
Recent findings indicate that complex cognitive functions are organized at a global level in
the brain and rely on large-scale information processing requiring functional integration of
multiple disparate neural assemblies. The critical question of the integration of distributed
brain activities is whether the essential integrative role can be attributed to a specific
structure in the brain or whether this ability is inherent to the cognitive network as a
whole. The results of the present study show that mean values of the running correlation
function in frontal-temporal EEG pairs with one electrode in the anterior cingulate cortex
(ACC) are significantly higher than the same values in other frontal-temporal pairs. These
findings indicate a particular role of the ACC in large-scale communication, which could
reflect its unique integrative functions in cognitive processing.
Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction
Today it is widely recognized that complex cognitive functions are organized at a global level in the brain, and that they
arise from more primitive functions organized in localized brain regions. The global mode of functioning relies on large-scale
information processing that requires mechanisms of functional integration of multiple disparate neural assemblies (Bressler
& Kelso, 2001; Fries, Reynolds, Rorie, & Desimone, 2001; Jensen, Kaiser, & Lachaux, 2007; Varela, Lachaux, Rodriguez, & Martinerie,
2001). The critical question of the integration of distributed brain activities is whether the essential integrative role
can be attributed to a specific structure in the brain or whether this ability is inherent to the cognitive network as a whole.
Recent conceptions concerning the integrative role of the anterior cingulate cortex (ACC) seem to support the first possibility,
which emphasizes structural localization of integrative functions (Paus, 2001; Posner, Rothbart, Sheese, & Tang, 2007; van
Veen & Carter, 2002). In principle, these concepts are supported by findings of anatomical connectivity of the ACC and its
structural particularities, especially the presence of spindle-shaped neurons with their widespread connections and their
putative role in long-range activity coordination, and by data obtained by neuroimaging methods during the performance
of cognitive tasks (Allman, Hakeem, Erwin, Nimchinsky, & Hof, 2001; Paus, 2001; van Veen & Carter, 2002). Numerous experimental
findings demonstrated the involvement of this structure in functions central to intelligent behavior, such as problem-solving,
error recognition, conflict detection and resolution, and adaptive responses to changing conditions (Allman
et al., 2001; Posner et al., 2007). This experimental evidence indicates that the role of the ACC in behavioral control includes
three main issues, i.e. its involvement in motor control, its proposed role in cognition, and its relationship to the motivation
states of the organism. On the basis of synthesis of available data, Paus (2001) proposed a hypothesis that the functional
1053-8100/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.concog.2010.04.007
* Corresponding author at: Department of Physiology, Faculty of Medicine, Masaryk University, Komenskeho 2, 66243 Brno, Czech Republic.
E-mail address: mkukleta@med.muni.cz (M. Kukleta).
Consciousness and Cognition 19 (2010) 879–886
Contents lists available at ScienceDirect
Consciousness and Cognition
journal homepage: www.elsevier.com/locate/concog
overlap of these three domains distinguishes the ACC from other frontocortical regions and that this overlap provides the
ACC with the potential to translate intentions into actions. A similar conceptual approach was developed by Rueda and Posner
(Posner et al., 2007; Rueda, Posner, & Rothbart, 2004), who based their proposal of the ACC role on the results of studies
designed to examine the mechanisms of self-regulation. They proposed that the major contribution of the ACC to brain functions
relies on its ability to regulate information influx from the environment in order to avoid conflicting responses in
behavior (Rueda et al., 2004).
Following the idea about the particular role of ACC in integrating distributed brain activities, the aim of the present study
was to test directly a hypothesis that functional interactions of the ACC with other remote brain regions distinguished this
structure from other frontocortical regions. With this aim we investigated the level of EEG synchrony in the beta-2 frequency
band between frontal and temporal depth electrode recordings in three different states occurring during a visual oddball
experiment: (i) throughout the whole experiment, i.e. regardless of the fact that several distinct mental operations subserved
the experimental task; (ii) 1 s before the stimulus, i.e. during the period of heightened expectation of the stimulus presentation;
(iii) 1 s after the stimulus, i.e. during the period of stimulus processing. The choice of the beta-2 frequency band was
based on demonstration of frequent occurrence of coherent oscillations in this band (Kukleta, Brazdil, Roman, Bob, & Rektor,
2009). In data analysis we compared mean values of the running correlation function calculated in 41 frontal-temporal pairs
of EEG records, which had their frontal electrodes in the gyrus cinguli, with values obtained in 139 frontal-temporal pairs
with their frontal electrodes in other parts of the frontal lobe.
2. Methods
2.1. Subjects
Eight patients (5 males, 3 females; aged 25–45 years; mean 32.4 years; all with medically intractable epilepsies; all righthanded)
participated in the study. Standard MicroDeep multilead depth electrodes (DIXI) with a diameter of 0.8 mm, length
of each recording contact 2 mm, and intercontact intervals of 1.5 mm were used for invasive EEG monitoring. The orthogonal
electrodes were implanted in the frontal, temporal, and/or parietal lobes using the methodology by Talairach et al. (1967). In
2 patients, additional diagonal electrodes were inserted stereotactically into the amygdalohippocampal complex (via frontal
approach, passing through the basal ganglia in 1 patient, via occipital approach in 1 patient). The electrodes were placed
bilaterally in 6 patients and unilaterally in 2 patients. The recording contacts at the electrode (5–15) were always numbered
from the medial to lateral sites. Their positions were indicated in relation to the axes defined by the Talairach system (1967)
using the ‘x, y, z’ format where ‘x’ is lateral, millimeters to midline, positive right hemisphere, ‘y’ is anteroposterior, millimeters
to the AC (anterior commissure) line, positive anterior, and ‘z’ is vertical, millimeters to the AC/PC (posterior commissure)
line, positive up. The exact positions of the electrodes in the brain were verified using post-placement magnetic
resonance imaging (MRI) with electrodes in situ. In all cases, both the number and positioning of the exploring electrodes
were determined by a diagnostic procedure the aim of which was to localize the seizure origin prior to a surgical treatment.
The recordings from lesional structures and epileptogenic zones were not included into the experimental analysis. No patient
from the group examined had bilateral hippocampal sclerosis or bilateral temporal lobe epilepsy. All the patients
had normal or corrected-to-normal vision and their performance in the experimental task was very good (median of errors
1.3%, minimum 0%, maximum 4.8%). The analysis of error responses to both non-target and target stimuli was taken as a
demonstration that they really processed the stimuli according to the experimental instructions. Informed consent was obtained
from each patient prior to the experiment, and the study received an approval from the Ethical Committee of Masaryk
University.
2.2. Procedure
The patients were seated comfortably in a moderately lighted room with a monitor screen positioned approximately
100 cm in front of their eyes. During the examination they were asked to focus the gaze continuously on the point in the
center of the monitor screen and to respond, as quickly as possible, to a target stimulus (a yellow letter X on white background)
by pressing a microswitch button in the dominant hand and counting the number of these stimuli in their heads,
and to ignore frequent stimuli (a yellow letter O on white background). Both stimuli were displayed on the black screen,
subtended at a visual angle of 3°. Their duration was 200 ms. The interstimulus intervals varied randomly between 2 and
5 s, the ratio of target to frequent, non-target stimuli was 1:5. The mean duration of the whole experiment was 17.4 min
(minimum 12.8 min, maximum 20.7 min).
2.3. EEG recording
The EEG signal was recorded simultaneously from various intra-cerebral structures using 64-channel Brain Quick EEG
system (Micromed). The recordings were monopolar with respect to a reference electrode placed on the right processus mastoideus
in all the cases. The EEGs were amplified with a bandwidth of 0.1–40 Hz at a sampling rate of 128 Hz. ScopeWin software
was used for the data analysis, which included up to 44 channels recorded simultaneously.
880 M. Kukleta et al. / Consciousness and Cognition 19 (2010) 879–886
2.4. Data analysis
In the data analysis EEG records from 180 frontal-temporal pairs were used. Frontal records were obtained from 27 sites;
temporal records from 56 sites (see Table 1A). The pairs were homolateral in 114 cases (52 pairs in the left, 62 pairs in the
right hemisphere), and heterolateral in the remaining 66 cases. According to the location of the frontal recording site these
pairs were divided into four groups (ACC – anterior cingulate cortex, OFC – orbital frontal cortex, DLFC – dorsolateral frontal
cortex, and MFC – mesial frontal cortex). More precise information about the location of the frontal recording sites is given in
Fig. 1. As demonstrated in Table 1B, records from anterior cingulate sites were paired with approximately the same set of
temporal sites as records from other frontal locations. All the investigated recording sites exhibited an evoked response
to target and non-target stimuli. One record from an electrode or two records derived from remote contacts of an electrode
were taken for the analysis only. As a rule, the largest response from similar ones was selected. The offline processing of each
record started with its frequency decomposition and reduction to the beta-2 frequency band (25–35 Hz) via a digital band
pass filter. The procedure comprised a spectrum computation using the Fast Fourier Transform, zeroing all the spectral components
outside the selected frequency interval, and an inverse complex Fast Fourier Transform computation. Then, in each
frontal-temporal pair of such filtered records, the correlation coefficients were calculated using the technique of running correlation.
The expert’s recommendation for the length of the sliding window in the computation of successive correlation
coefficients (94 ms) took into account the sampling frequency and the frequency band used.
Three correlation indicators were derived from each of the resulting correlation curves: (1) mean r-value of 30 s segment
of the correlation curve (18 values representing the whole experiment were obtained from each curve); (2) mean prestimulus
r-value calculated from averaged 1 s segments which preceded the stimulus onset; (3) mean poststimulus r-value calculated
from averaged 1 s segments which followed the stimulus onset. The second and third indicators were derived from
averaged data. In this case, 5 s segments of the correlation curves were averaged using the stimulus onset as the trigger
(À2.5 and +2.5 from stimulus onset). Segments linked to non-target stimuli were used only. One advantage of this selection
was the fact that the data obtained were not contaminated by efferent actions linked with movement; the other advantage
was the greater number of available responses, which allowed calculating the average curves with a more favorable signal/
noise ratio. The mean number of averaged stimuli was 136 (minimum 90, maximum 165). Only artifact-free segments were
included into the processing of both raw and averaged data (the selection was based on visual inspection of the segments by
an experienced person).
3. Results
The sequence of correlation coefficients obtained by the running correlation technique from the 180 frontal-temporal
pairs of filtered EEG records represented the source of data for analysis in this study. A typical time course of these correlation
curves is shown in Fig. 2, section A. The curves consisted of irregular oscillations between the maximal and the minimal
r-values; maximal peaks frequently attained values higher than +0.9. For the demonstration of the main characteristics
of these curves we chose the mean correlation coefficient calculated in 18 consecutive 30 s segments of each curve (the ‘‘30 s
r-values”). In a total of 3240 measurements obtained from 180 curves, their mean was 0.18 ± 0.13 (median 0.14, minimum
À0.10, maximum +0.71) and their distribution was quasi-normal (skewness 0.6, kurtosis 0.4). The values of this indicator
obtained from four different brain areas of four patients (64 individual measurements) are presented in Fig. 3, left section.
These data illustrate the most conspicuous feature of the indicator, i.e. its high and relatively constant standard deviations. In
Table 1
Number and location of recording sites from which 180 frontal-temporal pairs were created (A) and the percentage of temporal recording sites paired with sites
in the anterior cingulate gyrus and sites in other frontal locations (B).
A
Structure Number of sites (left/right hemisphere) Number of subjects
ACC (anterior cingulate gyrus) 3/4 6
OC (gyri orbitales) 4/4 7
DLFC (gyrus frontalis superior, medius, inferior, gyrus praecentralis) 3/3 5
MFC (gyrus frontalis medialis, gyrus rectus) 2/4 4
LTC (gyrus temporalis superior, medialis and inferior) 17/10 8
BTC (gyrus fusiformis and parahippocampalis) 5/4 5
HIPP (hippocampus) 7/4 6
AMY (amygdala) 5/4 7
B
Frontal sites Number of pairs LTC (%) BTC (%) HIPP (%) AMY (%)
ACC 41 49 14 20 17
OC 60 50 15 18 17
DLFC 41 46 15 20 19
MFC 38 50 8 21 21
M. Kukleta et al. / Consciousness and Cognition 19 (2010) 879–886 881
a total of 3240 analyzed measurements their mean value was 0.52 ± 0.03. We thought that this steadily high variability of the
indicator reflected the oscillatory nature of the mechanisms underlying the correlation curves.
Fig. 4, upper section, presents the ‘‘30 s r-values” of groups created according to the location of frontal recording sites. As
is evident, the level of frontotemporal activity correlation was highest in pairs with frontal recording sites in the anterior
cingulate cortex.
Statistical evaluation (Main effects ANOVA) showed a significant main effect of the recording site position
(F(3, 3229) = 66.3; p < .001); the p-values of the post hoc Scheffé test of differences between the anterior cingulate group
(ACC) on the one hand and OFC, DLFC, and MFC groups on the other hand were lower than .001 in all three comparisons.
The statistical procedure used took into account a possible influence of the position of the 30 s segment in the time course
of the experiment, and individual differences of the mechanisms underlying the correlation results. A more detailed investigation
of these two factors was beyond the scope of the present study.
With respect to the final interpretation of the results, the finding of a high correlation between the ‘‘30 s r-values” and the
number of peaks on the correlation curves, which exceeded the level of +0.9, was particularly important (Spearman R 0.84;
Fig. 1. Projection of 27 depth frontal orthogonal electrodes on lateral (A) or mesial (B) hemispheric surfaces. All electrodes, either right or left, were
positioned in the same scheme (left convexity, right mesial surface), according to the standard Talairach’s stereotaxic coordinate system. Allocation of single
recording sites for one of the four groups is represented by identical geometric symbols.
882 M. Kukleta et al. / Consciousness and Cognition 19 (2010) 879–886
Fig. 2. An illustration of the correlation curves calculated from non-averaged, raw data (section A) and from averaged data (section B). In section A the
curves represent, taken from top to bottom, the position of non-target (smaller marks) and target (higher mark) stimuli, the whole-band EEG records from
the left anterior cingulate cortex (ACC), and from the left superior temporal gyrus (GTS) of patient 1, filtered derivatives of these two records in the
frequency band of 25–35 Hz, and the resulting correlation curve. In section B the curves represent the position of non-target stimuli, the averaged pre- and
poststimulus whole-band records from the left anterior cingulate cortex (ACC) and the left superior temporal gyrus (GTS) of patient 1, the filtered
derivatives of these two records in the frequency band of 25–35 Hz, and the resulting correlation curve. The vertical line passing through all curves marks
the stimulus onset.
M. Kukleta et al. / Consciousness and Cognition 19 (2010) 879–886 883
p < .001). This finding allowed considering the ‘‘30 s r-values” as a measure of interregional activity synchronization. The
number of such peaks in one 30 s segment varied from 2 to 73 (median 14) and the distribution of its values was not normal.
Mean r-values in the 1 s segments, which preceded and followed the non-target stimulus onset, represented another two
correlation indicators analyzed in the study. Both indicators varied largely in a total of 180 averaged correlation curves. The
mean prestimulus r-value was 0.19 ± 0.24 (median 0.19, minimum À0.37, maximum +0.78), the mean poststimulus r-value
was 0.25 ± 0.24 (median 0.25, minimum À0.35, maximum +0.81). The middle and bottom sections of Fig. 4 present mean
prestimulus and poststimulus r-values of groups created according to the location of frontal recording sites. Even these indicators
showed that the pairs with frontal recording sites in the cingulate cortex had higher mean r-values than pairs from
other frontal locations. ANOVA showed a significant main effect for the recording site position in both cases (F(3169) = 13.2
in the first and 8.0 in the second case; p < .001 in both cases). The p-values of the post hoc Scheffé test evaluating the significance
of differences between the anterior cingulate group (ACC) on the one hand and OFC, DLFC, and MFC groups on
the other hand were lower than .001, .001 and .012, respectively, in comparisons of prestimulus values, and .001, .002
and equal to .120, respectively, in comparisons of poststimulus values. In other words, the pairs with their frontal recording
sites in the cingulate cortex had higher mean r-values than had pairs from other frontal locations in five of six comparisons.
A comparison of eleven ACC pairs located in the right hemisphere and of eleven ACC pairs in the left hemisphere revealed
a significant main effect of hemisphere location in the case of the ‘‘30 s r-values” (ANOVA, F(1372) = 5.9, p = .016; means
0.26 ± 0.10 in the right and 0.13 ± 0.08 in the left hemisphere). Prestimulus and poststimulus ACC r-values in the right
and the left hemisphere did not differ significantly (ANOVA, F(1, 15) = 0.7 and 2.7, respectively, p > .05 in both cases). Due
to the limited number pairs and inconsistency of the result, the difference between right and left locations of the recording
site pairs was not analyzed in greater detail.
Fig. 3. An illustration of mean values of the correlation coefficient and their standard deviation calculated from four different brain areas of four patients in
eighteen successive 30 s segments of raw correlation curves (left section) and in 1 s prestimulus and poststimulus epochs of averaged correlation curves
(right section). The frontal recording site was in the anterior cingulate cortex of patient 1 (ACC), in the orbital frontal cortex of patient 2 (OFC), in the
dorsolateral frontal cortex of patient 3 (DLFC), and in the mesial frontal cortex of patient 4 (MFC).
884 M. Kukleta et al. / Consciousness and Cognition 19 (2010) 879–886
4. Discussion
The results of the present study support recent findings suggesting a specific integrative role of the ACC (Paus, 2001; Posner
et al., 2007). Particularly, these results show that mean values of the running correlation function in frontal-temporal
EEG pairs with one recording contact in the cingulate cortex (ACC pairs) are significantly higher than mean values of running
correlation in other frontal-temporal pairs. This regional difference in the phase synchrony was found in three distinct functional
states: (i) throughout the whole experiment; (ii) during the period of heightened expectation of the next stimulus; (iii)
during the cognitive elaboration of the stimulus. The finding suggests a state-independent stronger communication of the
ACC with other components of the cognitive network. A crucial interpretation of this finding is that higher synchrony in
the ACC pairs in comparison with other frontal-temporal pairs may represent a particular role of the ACC in large-scale communication,
which could reflect its unique integrative functions in cognitive processing. This interpretation could have key
consequences for the understanding of the neural correlate of consciousness, which although it is spatially distributed and
related to large-scale integration may have its ‘‘extraordinary places” with a specific integrative role.
The majority of recent studies on integrative or binding mechanisms have focused on EEG analysis and observed functionally
relevant epochs of synchronization mainly in the gamma frequency band in various species and brain structures during
attention, perception, motor and memory tasks (Jensen et al., 2007; Lee, Williams, Breakspear, & Gordon, 2003; Singer,
2001). A crucial result of these studies, which presents a direct link between visual perception and gamma synchrony,
was reported by Eckhorn, Singer and colleagues in the cat visual cortex (Eckhorn et al., 1988). Following this finding, the
functional significance of synchronous gamma activity in selective attention, perceptual processing, and recognition was
repeatedly demonstrated in animal and human studies (Fries et al., 2001; Jensen et al., 2007; Meador, Ray, Echauz, Loring,
& Vachtsevanos, 2005; Rodriguez, Kallenbach, Singer, & Munk, 2004). The interpretation of all these findings seems to imply
a common assumption that transient and precise synchronization of neuronal discharges represents a crucial binding mechanism
and is related to the emergence of conscious mental states. The concept of dynamic binding by synchronization of
Fig. 4. Mean r-values in groups of correlated frontotemporal pairs created according to the location of their frontal recording site (ACC – anterior cingulate
cortex; OFC – orbital frontal cortex; DLFC – dorsolateral frontal cortex; MFC – mesial frontal cortex). The upper section presents the means from the 30 s
segments of correlation curves (N 738, 1080, 738, and 684, respectively), the middle and bottom sections present the means from the 1 s prestimulus and
poststimulus epochs (N 41, 60, 41, and 38, respectively, in both indicators). Vertical bars denote 0.95 confidence intervals.
M. Kukleta et al. / Consciousness and Cognition 19 (2010) 879–886 885
distributed neuronal activity has been developed mainly in the context of perceptual processing. With respect to visual
awareness, the available evidence shows that singular object features such as color, shape, texture, size, brightness, etc. produce
activity, at least to some degree, in separate areas of the visual cortex (Bressler & Kelso, 2001; Crick & Koch, 1992; Singer,
2001). Recent neuroscience, however, has not located a distinct place in the brain, in which distributed visual information
comes together (Bartels & Zeki, 2006; Crick & Koch, 1992). Additionally, there is evidence that the concept of binding is useful
to apply to other domains of brain functioning such as attention, memory formation and recall, motor control, sensorimotor
integration, language processing, and logical inference (Singer, 1993, 2001). Following these findings the large
majority of authors share the view that consciousness results from a cooperative process in a highly distributed network,
and is not attributable to a single brain structure or process. Instead of a single central place (‘‘Cartesian theatre”), there
are various events of content fixation that occur in various places at various times in the brain (Dennett, 1991, pp. 365).
The evidence for this view of consciousness is presented by a whole series of experimental results in cognitive neuroscience
and psychology (van der Velde & de Kamps, 2006; Varela et al., 2001; Zeki, 2003).
According to this view, the neural correlate of consciousness represents a part of the nervous system that transforms neural
activity in reportable subjective experiences. A major hypothesis concerning its functioning is that the network can compare
and bind activity patterns only if they arrive at the system simultaneously (van der Grind, 2002). Consciousness
combines the present multimodal sensory information with relevant elements of the past and creates spatiotemporal memory.
Information from each modality is continuously distributed into distinct features and locally processed in different relatively
specialized brain regions and globally integrated by interactions among these regions. The resulting subjective
experience is then represented by integration through levels of synchronization within neuronal populations and by
large-scale integration among multiple brain regions (Bob, 2009; John, 2002; King, 2009; Singer, 2001).
Within this context the results of this study present the first reported evidence that the level of binding between structures
could be significantly spatially differentiated and, although the synchronization underlying cognitive processes presents
a global phenomenon related to large-scale integration, there are ‘‘extraordinary places” within the brain with a
specific integrative role.
Acknowledgments
The authors thank for support by Grants MSM 0021620849 and MSM 0021622404.
References
Allman, J. M., Hakeem, A., Erwin, J. M., Nimchinsky, E., & Hof, P. (2001). The anterior cingulate cortex. The evolution of an interface between emotion and
cognition. Annals of the New York Academy of Sciences, 935, 107–117.
Bartels, A., & Zeki, S. (2006). The temporal order of binding visual attributes. Vision Research, 46, 2280–2286.
Bob, P. (2009). Quantum science and the nature of mind. Journal of Mind and Behavior, 30, 1–14.
Bressler, S. L., & Kelso, J. A. S. (2001). Cortical coordination dynamics and cognition. Trends in Cognitive Sciences, 5, 26–36.
Crick, F., & Koch, Ch. (1992). The problem of consciousness. Scientific American, 267, 153–159.
Dennett, D. (1991). Consciousness explained. Little, Brown: Boston.
Eckhorn, R., Bauer, R., Jordan, W., Brosch, M., Kruse, W., Munk, M., et al (1988). Coherent oscillations: A mechanism of feature linking in the visual cortex?
Multiple electrode and correlation analysis in the cat. Biological Cybernetics, 60, 121–130.
Fries, P., Reynolds, J. H., Rorie, A. E., & Desimone, R. (2001). Modulation of oscillatory neuronal synchronization by selective visual attention. Science, 291,
1560–1563.
Jensen, O., Kaiser, J., & Lachaux, J. P. (2007). Human gamma-frequency oscillations associated with attention and memory. Trends in Neurosciences, 30,
317–324.
John, E. R. (2002). The neurophysics of consciousness. Brain Research Reviews, 39, 1–28.
King, C. (2009). Sensory transduction and subjective experience: Expression of eight genes in three senses suggests a radical model of consciousness.
Activitas Nervosa Superior, 51, 45–50.
Kukleta, M., Brazdil, M., Roman, R., Bob, P., & Rektor, I. (2009). Cognitive network interactions and beta 2 coherence in processing non-target stimuli in visual
oddball task. Physiological Research, 58, 139–148.
Lee, K. H., Williams, L. M., Breakspear, M., & Gordon, E. (2003). Synchronous gamma activity: A review and contribution to an integrative neuroscience
model of schizophrenia. Brain Research Reviews, 41, 57–78.
Meador, K. J., Ray, P. G., Echauz, J. R., Loring, D. W., & Vachtsevanos, G. J. (2005). Gamma coherence and conscious perception. Neurology, 59, 1847–1854.
Paus, T. (2001). Primate anterior cingulate cortex: Where motor control, drive and cognition interface. Nature Reviews Neuroscience, 2, 417–424.
Posner, M. I., Rothbart, M. K., Sheese, B. E., & Tang, Y. (2007). The anterior cingulate gyrus and the mechanism of self-regulation. Cognitive, Affective, and
Behavioral Neuroscience, 7, 391–395.
Rodriguez, R., Kallenbach, U., Singer, W., & Munk, M. H. (2004). Short- and long-term effects of cholinergic modulation on gamma oscillations and response
synchronization in the visual cortex. Journal of Neuroscience, 24, 10369–10378.
Rueda, M. R., Posner, M. I., & Rothbart, M. K. (2004). Attentional control and self-regulation. In R. F. Baumeister & K. D. Vohs (Eds.), Handbook of self-regulation
research theory and applications (pp. 283–300). New York: Guilford.
Singer, W. (1993). Synchronization of cortical activity and its putative role in information processing and learning. Annual Review of Physiology, 55, 349–374.
Singer, W. (2001). Consciousness and the binding problem. Annals of the New York Academy of Sciences.
Talairach, J., Szikla, G., Tournoux, P., Prosalentis, A., Bordas-Ferrer, A., Covello, J., et al (1967). Atlas d’Anatomie Stéréotaxique du Telencephale. Masson.
van der Grind, W. (2002). Physical, neural, and mental timing. Consciousness and Cognition, 11, 241–264.
van der Velde, F., & de Kamps, M. (2006). Neural blackboard architectures of combinatorial structures in cognition. Behavioral and Brain Sciences, 29, 1–72.
van Veen, V., & Carter, C. S. (2002). The anterior cingulate as a conflict monitor: fMRI and ERP studies. Physiology & Behavior, 77, 477–482.
Varela, F. J., Lachaux, J. P., Rodriguez, E., & Martinerie, J. (2001). The brainweb: Phase synchronization and large-scale integration. Nature Reviews
Neuroscience, 2, 229–239.
Zeki, S. (2003). The disunity of consciousness. Trends in Cognitive Neuroscience, 7, 214–218.
886 M. Kukleta et al. / Consciousness and Cognition 19 (2010) 879–886
Příloha č. 8
Brazdil M, Babiloni C, Roman R, Daniel P, Bares M, Rektor I, Eusebi F, Rossini PM, Vecchio
F. Directional Functional Coupling of Cerebral Rhythms Between Anterior Cingulate and
Dorsolateral Prefrontal Areas During Rare Stimuli: A Directed Transfer Function Analysis of
Human Depth EEG Signal. Human Brain Mapping 2009;30:138-146. IF (2009) = 6,256
Directional Functional Coupling of Cerebral
Rhythms Between Anterior Cingulate and
Dorsolateral Prefrontal Areas During Rare Stimuli:
A Directed Transfer Function Analysis of
Human Depth EEG Signal
Milan Bra´zdil,1
Claudio Babiloni,2,3,4
Robert Roman,5
Pavel Daniel,1
Martin Baresˇ,1
Ivan Rektor,1
Fabrizio Eusebi,2,6,7
Paolo Maria Rossini,4,8,9
and Fabrizio Vecchio2,4
*
1
First Department of Neurology, Masaryk University, St. Anne’s Hospital, Brno, Czech Republic
2
Dipartimento di Fisiologia Umana e Farmacologia, Universita` ‘‘La Sapienza,’’ Rome, Italy
3
Casa di Cura San Raffaele Cassino, San Raffaele Pisana, Italy
4
A.Fa.R Dip. Neuroscienze - Ospedale FBF; Isola Tiberina, Rome, Italy
5
Department of Physiology, Masaryk University, Brno, Czech Republic
6
Istituto di Medicina e Scienza dello Sport, CONI Servizi, Rome, Italy
7
IRCCS Neuromed, Pozzilli (IS), Italy
8
Clinica Neurologica, Universita` ‘‘Campus Biomedico,’’ Rome, Italy
9
IRCCS ‘‘Centro S. Giovanni di Dio-FBF,’’ Brescia, Italy
Abstract: What is the neural substrate of our capability to properly react to changes in the environment? It
can be hypothesized that the anterior cingulate cortex (ACC) manages repetitive stimuli in routine conditions
and alerts the dorsolateral prefrontal cortex (PFC) when stimulation unexpectedly changes. To provide
evidence in favor of this hypothesis, intracerebral stereoelectroencephalographic (SEEG) data were
recorded from the anterior cingulate and dorsolateral PFC of eight epileptic patients in a standard visual
oddball task during presurgical monitoring. Two types of stimuli (200 ms duration) such as the letters O
(frequent stimuli; 80% of probability) and X (rare stimuli) were presented in random order, with an interstimulus
interval between 2 and 5 s. Subjects had to mentally count the rare (target) stimuli and to press a button
with their dominant hand as quickly and accurately as possible. EEG frequency bands of interest were y
(4–8 Hz), a (8–12 Hz), b (14–30 Hz), and g (30–45 Hz). The directionality of the information flux within the
EEG rhythms was indexed by a directed transfer function (DTF). The results showed that compared with
the frequent stimuli, the target stimuli induced a statistically significant increase of DTF values from the
anterior cingulate to the dorsolateral PFC at the y rhythms (P < 0.01). These results provide support to the
hypothesis that ACC directly or indirectly affects the oscillatory activity of dorsolateral PFC by a selective
frequency code under typical oddball conditions. Hum Brain Mapp 30:138–146, 2009. VVC 2007 Wiley-Liss, Inc.
Key words: connectivity; information flow; oddball paradigm; directed transfer function method (DTF);
prefrontal cortex; stereoEEG
Contract grant sponsor: MSˇMT Cˇ R Research; Contract grant number:
MSM0021622404.
*Correspondence to: Dr. Fabrizio Vecchio, Dipartimento di Fisiologia
Umana e Farmacologia, Universita` degli Studi di Roma ‘‘La
Sapienza,’’ P.le Aldo Moro 5, 00185 Rome, Italy.
E-mail: fabrizio.vecchio@uniroma1.it
Received for publication 3 July 2007; Revised 28 August 2007;
Accepted 29 August 2007
DOI: 10.1002/hbm.20491
Published online 12 November 2007 in Wiley InterScience (www.
interscience.wiley.com).
VVC 2007 Wiley-Liss, Inc.
r Human Brain Mapping 30:138–146 (2009) r
INTRODUCTION
One of the most widely utilized experimental paradigms
in cognitive neuroscience is the oddball task, which has
been used extensively to study target rare stimulus processing
in the human brain. It is a simple discrimination
task with a randomly alternating presentation of two types
of sensory stimuli: one frequent and one rare. The subject
performing the oddball task is instructed to ignore frequent
stimuli and to perform a certain task (motor
response, counting, etc.) when rare (target) stimuli are
detected. Target stimuli in this task indubitably trigger in
the brain a whole set of cognitive processes that are mostly
linked to attentional mechanisms, but concurrently they
also reflect other functions collectively called ‘‘executive’’
including short term memory, assessment of stimulus relevance,
decision making, and response. It is noteworthy
that most of these processes are not triggered by frequent
stimuli. Interestingly, a P3 component of event-related
potentials, evoked by rare (target) stimuli in the subject’s
averaged electroencephalogram, can be subdivided into
two subcomponents. A frontal P3a is enhanced by nontarget
rare and by novel (distractor) stimuli in an extended
oddball task (three types of stimuli—target, standard, and
distractor). A later parietal P3b is enhanced by the certainty
of target detection [Snyder and Hillyard, 1976;
Squires et al., 1975]. The P3a seems to reflect the attentional
functions, while P3b has been suggested to embody
the closure of the cognitive event-encoding cycle or to
reflect a whole set of processes that mediate between perceptual
analysis and response initiation [Halgren et al.,
1998; Verleger et al., 2005].
Among the attention-related brain regions under study,
two frontal regions, the anterior cingulate cortex (ACC)
and the dorsolateral prefrontal cortex (PFC), appear to be
particularly involved in the attentional processes induced
by unpredictable changes in the environment and also in a
relative selection of proper reactions. These areas are also
unequivocal generators of P3a potential, as revealed
repeatedly by intracranial event-related potential studies
[Baudena et al., 1995; Bra´zdil et al., 1999; Halgren et al.,
1998]. To capture their strict functional relationship, it has
been speculated that the ACC and the PFC cooperate to
regulate behavior. The ACC would be specifically responsible
for monitoring stimulus processing/responding and
for signaling to the PFC and general arousal systems when
extra-activation is needed. The PFC would be specifically
responsible for selectively enhancing task-related processes
and stimulus representations in posterior cortical areas, to
meet endogenous plans and instructions [Banich et al., 2000;
Barch et al., 2001; Botvinick et al., 1999; Braver et al., 2001;
Carter et al., 1998; Cohen et al., 2000; Luks et al., 2002]. In
this sense, the ACC may be considered as a key area in
the tuning of earlier stages of task-driven allocation of
attention prior to stimulus processing [Gitelman et al.,
1999; Woldorf et al., 2001]. It can be maintained that the
ACC plays a function quite similar to that of a ‘‘supervisory
attentional system’’ [Shallice, 1988; Shallice et al.,
1989], which would manage repetitive stimulation in routine
conditions and would alert the dorsolateral PFC
when the environment unexpectedly changes and proper
cognitive processes and reactions have to be set [Luks
et al., 2002; MacDonald et al., 2000].
Studies on functional integration between the ACC and
the dorsolateral PFC have described how these functionally
specialized areas interact and how these interactions
depend on changes of context. In a recent fMRI study of
effective connectivity during a visual oddball task [Bra´zdil
et al., 2007], a simple hierarchy within the right frontal
lobe has been revealed, with the ACC exerting influence
over the PFC. This finding has supported the NormanShallice
model of action control [Shallice, 1988] with the
presumed central role of the ACC.
Keeping in mind the above results [Bra´zdil et al., 2007],
one may argue that due to its low temporal resolution,
fMRI cannot provide definitive evidence on the directionality
of functional connectivity between cortical areas during
quick stimulus processing, especially if the functional
coupling of brain rhythms is implied in that connectivity.
The direction of the cortico–cortical information flows can
be studied by the computation of directed transformation
function (DTF) from electroencephalographic (EEG) rhythms
[Kaminski and Blinowska, 1991]. With the DTF technique, it
has recently been shown that the presleep period is characterized
by a parietooccipital-to-frontal cortical information
flow, while the opposite holds at sleep onset [De Gennaro
et al., 2004]. It has been also demonstrated that fronto-parietal
directional flows varied during memory processes [Babiloni
et al., 2004, 2006]. During mere encoding of visual information,
the parietooccipital-to-frontal direction of the cortical
information flow predominates. The direction of fronto-parietal
fluxes is balanced during short- or long (retrieval)-memory
processes [Babiloni et al., 2004, 2006]. However, to our
knowledge, DTF technique has been used neither for the
investigation of attentional functions during oddball tasks
nor for the estimation of interactions between the ACC and
the dorsolateral PFC after rare stimuli.
The aim of the present study was to investigate the effective
functional connectivity between the ACC and the dorsolateral
PFC during task-relevant processing of rare events
(visual oddball paradigm). The DTF method was applied to
human intracerebral EEG data recorded from the ACC and
the PFC. The directionality of the information flow between
these two regions of interest was estimated from SEEG
rhythms, in order to evaluate the working hypothesis that
the ACC directly or indirectly affects the oscillatory activity
of the PFC under oddball conditions.
METHODS AND MATERIALS
Subjects
Eight patients (six males and two females) ranging in age
from 19 to 37 years (with an average age of 27.9 6 1.9 SE),
r Effective Connectivity in Rare Stimuli Processing r
r 139 r
all with medically intractable epilepsies, participated in
the study. Depth electrodes were implanted to localize the
seizure origin before surgical treatment. The intracerebral
electrodes were implanted orthogonally under stereotaxic
conditions (stereoEEG, SEEG) using the methodology of
Talairach et al. [1967]. Standard MicroDeep semiflexible electrodes
(DIXI) with a diameter of 0.8 mm, a length of each
contact of 2 mm, and an intercontact interval of 1.5 mm
were used for invasive EEG monitoring. The exact positions
of the electrode contacts in the brain were verified using
postplacement MRI with electrodes in situ (see Fig. 1).
Lesional anatomical structures and epileptogenic zone structures
were not included in the analysis. All subjects had normal
or corrected-to-normal vision. All the patients were able
to fully understand and perform the experimental task.
Informed consent was obtained from each subject before the
investigation and the study received approval from the
Ethics Committee of Masaryk University Brno.
Visual Oddball Task
Subjects were seated comfortably in a moderately
lighted room with a monitor screen positioned $100 cm in
front of their eyes. During the examination, they were
requested to continuously focus their eyes on the small fixation
point in the center of the screen and to minimize
blinking. A standard visual oddball task was performed;
two types of stimuli (frequent and rare) were presented in
the center of the screen in random order. Clearly visible
yellow capital letters O (frequent) and X (rare; $50 trials)
on a white background were used. The duration of stimuli
exposure was constant at 200 ms; the ratio of rare to frequent
stimuli was 1:5. The interstimulus interval randomly
varied between 2 and 5 s. Each subject was instructed to
respond to the rare (target) stimulus as quickly and accurately
as possible by pressing a microswitch button in the
dominant hand, and at the same time was instructed to
mentally count the target stimuli.
EEG Recordings
The EEG signal was simultaneously recorded from various
intracerebral structures and from the CPz scalp electrode
(situated between Cz and Pz) using the 64-channel
Brain Quick EEG system (Micromed). All recordings were
monopolar with respect to a reference electrode on the
processus mastoideus. All impedances were less then
5 kX. EEGs were amplified with a bandwidth of 0.1–40 Hz
at a sampling rate of 128 Hz. Further processing was performed
with artifact-free EEG periods. Eye movement artifacts
were monitored via an electrooculogram (EOG)
recorded by electrodes at the lateral canthus of the eyes.
Surface electromyographic activity of bilateral extensor
digitorum muscles was also recorded to monitor the voluntary
movements as well as involuntary mirror movements
and muscle activations.
In all subjects, we considered the depth electrodes crossing
the dorsolateral PFC (Brodmann areas—BAs 9, 44, 45,
46) and ACC (BAs 24 and 32). To evaluate the effect of the
rare stimulus independently by hemisphere, we considered
the right hemisphere for half of the subjects, and the
left hemisphere for the other half.
The collected SEEG data were segmented in single trials.
For both rare (target) and frequent stimuli, the data windows
for the spectral analysis were two 0.5 s windows: pre (a period
of 0.5 s just before the stimulus) and post (a period of
0.5 s after the stimulus). The SEEG segments showing motor
or instrumental artifacts were rejected. Eight SEEG recording
sites from the PFC and eight SEEG recording sites from the
ACC were considered for the data analysis (see Fig. 2).
‘‘Direction’’ of the Functional Connectivity
Estimated by the MVAR Model
Before computing the DTF, the SEEG data were preliminarily
normalized by subtracting the mean value and by
dividing for the variance, according to standardized rules
by Kaminski and Blinowska [1991]. The DTF can be
Figure 1.
Position of orthogonal depth electrode
[passing through PFC into ACC]. The real
volume of the electrode is about 10% of
the displayed artifact.
r Bra´zdil et al. r
r 140 r
considered as a normalized value ranging from 0 to 1. An
important step of the DTF method was the computation of
the so-called MVAR model [Blinowska et al., 2004; Kaminski
and Blinowska, 1991; Kaminski et al., 1997]. SEEG data
at all the electrode contacts inside the two regions of interest
(anterior cingulate and dorsolateral prefrontal cortices)
were simultaneously given as an input to the MVAR
model for the computation of the DTF among all combinations
of electrode pairs. The DTF values were obtained as
the mean between the direction among each electrode contact
of the ACC and all the contacts of the dorsolateral
PFC. This procedure guaranteed the comparability of the
results across subjects. The same procedure was true for
the evaluation of the opposite direction of the DTF values
between the two cortical regions of interest. In nonmathematical
terms, the MVAR model estimates information
flow between recording sites A and B by computing the
extent to which the EEG data at recording sites A can be
predicted based on the EEG data of recording sites B and
vice-versa. A direction of the information flow from A to B
is stated when that case is statistically more probable than
directionality from B to A.
The mathematical core of the MVAR algorithm is based
on the ARfit programs running on the platform (Matlab 5.3,
MathWorks, Natick, MA). The model order was 7, as estimated
by the Akaike criterion suggested in previous DTF
studies [Kaminski and Blinowska, 1991; Kaminski et al.,
1997]. The goodness of fit was evaluated by visual inspection
of the values of noise matrix V of the MVAR model.
The MVAR model is defined as
Xp
j¼0
AjXtÀj ¼ Et
where Xt is the L-dimensional vector representing the Lchannel
signal at time t; Et is white noise; Aj is the L 3 L
matrix of the model coefficients; and p is the number of
time points considered in the model. From the identified
coefficients of the model Aj, spectral properties of the signals
can be obtained by the following z-transformation of
the above equation:
XðzÞ ¼ HðzÞEðzÞ
where H(z) is a transfer function of the system and
HðzÞ ¼
Xp
j¼0
AjZÀj
0
@
1
A
À1
zÀi
¼ expðÀi2pfdtÞ
Since the transfer function H( f ) is not a symmetric matrix,
the information transmission from the jth to the ith
channel is different from that from the ith to the jth channel.
The DTF from the jth channel to the ith channel is
defined as the square of the element of H( f ) divided by
the squared sum of all elements of the relevant row.
DTFijðfÞ ¼
Hij



2
PL
m¼1 HimðfÞj j2
A substantial difference between DTF( f )ij and DTF( f )ji
may suggest an asymmetric information flow from electrode
i to electrode j. When DTF( f )ij is greater in magnitude
than DTF( f )ji, the ‘‘direction’’ of the information flow
is from electrode j to electrode i. On the other hand, the
‘‘direction’’ of the information flow is from electrode i to
electrode j, when DTF( f )ji is greater in magnitude than
DTF( f )ij. Noteworthy, the present procedure is based on
the assumption that AR model is quasi stationary over the
500 ms window used for the MVAR analysis. This
assumption is reasonable for short period of the brain activity
and is the basis of the countless literature commonly
Figure 2.
Schematic distribution of recording
sites within the lateral
prefrontal cortex (left) and the
anterior cingulate (right). The
recording sites of each subject
are labeled with specific symbols;
each subject has its own
distinguishing symbol.
r Effective Connectivity in Rare Stimuli Processing r
r 141 r
using fast Fourier transform, spectral coherence, and DTF
measurements in event-related paradigms.
The DTF values were evaluated as the maximum DTF
peak value at each of the following EEG bands of interest:
y (4–8 Hz), a (8–12 Hz), b (14–30 Hz), and g (30–45 Hz).
The DTF values were computed for the frequent and rare/
target conditions and for the pre and poststimulus periods
defined as the period of 0.5 s before and 0.5 s after the
onset of the frequent or rare/target stimulus.
Statistical Analysis
A preliminary Kolgomorov-Smirnoff test evaluated the
Gaussian distribution of the DTF data The results showed
that all the variables were Gaussian so that DTF data were
given as an input to repeated measures ANOVA. The
Mauchley test evaluated the sphericity assumption and
correction of the degrees of freedom was carried out using
the Greenhouse-Geisser procedure. The Duncan test was
used for post-hoc comparisons (P < 0.05). The side of the
electrode position (namely left or right hemisphere) was
used as covariate. Because of the small number of the subjects,
no statistical analysis could address the effects of
hemispherical lateralization or those the activity within
subregions of prefrontal (Brodmann areas—BAs 9, 44, 45,
46) and ACC (BAs 24 and 32).
To simplify the visualization and statistical analysis of the
DTF results, the directional flow of information was given
as ‘‘anterior cingulate-to-dorsolateral PFC’’ minus ‘‘dorsolateral
prefrontal-to-ACC’’ direction of the DTF. Positive values
of this subtraction (DTFdiff values higher than zero)
showed the predominance of the ‘‘anterior cingulate-to-dorsolateral
PFC’’ direction over the ‘‘dorsolateral prefrontal-toACC’’
direction of the DTF. On the other hand, negative values
of this subtraction (DTFdiff values lower than zero)
showed the predominance of the ‘‘dorsolateral prefrontal-toACC’’
direction over the ‘‘anterior cingulate-to-dorsolateral
PFC’’ direction of the DTF. We performed four different
ANOVA analyses using DTFDiff values as a dependent
variable; each ANOVA regarded a single band of interest (y,
a, b, and g). The ANOVA factors were Condition (frequent,
target) and Period (prestimulus, poststimulus).
RESULTS
Behavioral Results
Satisfactory co-operation of all subjects was observed
during the experiment. The mean reaction time in the
group of patients was 502 ms (6SD 31.3). The mean counting
accuracy (i.e., accuracy of subjects’ reports at the end
of the experiment) was 88.6% (6SD 11.03).
Electrophysiological Results
Figure 3 illustrates the across-subjects mean DTF values
(1–48 Hz) relative to the ‘‘anterior cingulate-to-dorsolateral
PFC’’ direction and the ‘‘dorsolateral prefrontal-to-ACC’’
direction for the frequent and target trials in both pre and
poststimulus periods. Similar DTF values are noted relative
to the ‘‘anterior cingulate-to-dorsolateral PFC’’ and the
‘‘dorsolateral prefrontal-to-ACC’’ directions in the prestimulus
periods of both frequent and target stimuli as well
as in the poststimulus period of the frequent stimuli. With
respect to these DTF values, there is a clear increment of
the DTF values relative to the ‘‘anterior cingulate-to-dorsolateral
PFC’’ in the poststimulus period of the target stimuli,
especially at low EEG frequencies. These DTF values
Figure 3.
Grand average (N 5 8) of the DTF values (1–48 Hz) relative to
the ‘‘anterior cingulate-to-dorsolateral prefrontal cortex’’ direction
and the ‘‘dorsolateral prefrontal-to-anterior cingulate cortex’’
direction for the frequent and target trials in both pre- and
post-stimulus periods. [Color figure can be viewed in the online
issue, which is available at www.interscience.wiley.com.]
Figure 4.
Mean DTFdiff values (6standard error, SE) relative to the bands
of interest (y, a, b, and g). Of note, positive DTFdiff values illustrate
the predominance of the ‘‘anterior cingulate-to-dorsolateral
prefrontal cortex’’ direction over the ‘‘dorsolateral prefrontal-toanterior
cingulate cortex’’ direction of the DTF; the opposite is
true for the negative DTFdiff values. A statistical significance of the
changes is depicted by asterisks (* P < 0.0292; ** P 5 0.006).
[Color figure can be viewed in the online issue, which is available
at www.interscience.wiley.com.]
r Bra´zdil et al. r
r 142 r
were used as an input for the computation of the DTFDiff
values. Figure 4 plots mean DTFdiff values (6standard
error, SE) relative to the bands of interest (y, a, b, and g).
Of note, positive DTFdiff values illustrate the predominance
of the ‘‘anterior cingulate-to-dorsolateral PFC’’ direction
over the ‘‘dorsolateral prefrontal-to-ACC’’ direction of
the DTF; the opposite is true for the negative DTFdiff values.
In line with the results relative to the DTF values,
DTFdiff values point to the increment of the directionality
‘‘anterior cingulate-to-dorsolateral PFC’’ in the poststimulus
period of the target stimuli at a, b, and y bands.
The ANOVA analyses of the DTFDiff values showed a
statistically significant interaction between the Condition
(frequent, target) and Period (prestimulus, poststimulus)
factors only for the y band (F(1,7) 5 7.47; P < 0.0292). Post
hoc analysis disclosed a statistically significant increment
of the positive values of the DTFdiff in the poststimulus
period of the target stimuli when compared with the prestimulus
period (P 5 0.006), indicating a global prevalence
of ‘‘anterior cingulate-to-dorsolateral PFC’’ direction over
‘‘dorsolateral prefrontal-to-ACC’’ direction of the DTF. Furthermore,
there was a statistically significant increment of
the positive values of the DTFdiff in the poststimulus period
of the target stimuli in comparison to the frequent
stimuli (P 5 0.01), confirming the specificity of the result
mentioned earlier.
One may argue that the above results merely reflected
the anticipation of P3 in the ACC with respect to the dorsolateral
PFC. To address this issue, we averaged SEEG
trials to form the P3 in the ACC and dorsolateral PFC
[for methodology see Bra´zdil et al., 1999]. The latency of
the P3 was computed for the two regions of interest
(ROI) and compared. Figure 5 shows the P3 in these
regions for two representative subjects. If independent
local generators of P3 potentials were observed in both
ACC and PFC (which was the situation in three cases),
either identical P3 latencies were found in both ROI or a
shorter latency of P3 was even found in PFC (in one subject).
In five subjects, typical P3 sources have been proven
just in one recording site (ACC or PFC), whilst
obvious far field potentials were observed in the other
ROI (PFC in 2 cases; ACC in one case) or P3 was even
completely missing there (PFC in two cases). In the majority
of records, the comparison of the P3 latencies therefore
could not be justified. From this simple analysis, it
can be seen that the P3 peak was not earlier in latency in
the ACC when compared with dorsolateral PFC. It is therefore
unlikely that our DTF results are due to differences in
the P3 latency between ACC and dorsolateral PFC.
DISCUSSION
Does the ACC affect the oscillatory activity of dorsolateral
PFC activity, to properly react to unpredictable
changes in the environment? To explore this issue, we
proposed an advanced methodological approach applying
the DTF analysis of directional information fluxes within
Figure 5.
Intracerebral event-related potentials
from two of our subjects (A
and B) recorded from ACC and
dorsolateral PFC [for methodology
see Bra´zdil et al., 1999].
Note exactly the same latency of
P3 potential in mesial and lateral
aspect of the frontal lobe in subject
A, and somewhat earlier P3
latency within PFC in subject B.
r Effective Connectivity in Rare Stimuli Processing r
r 143 r
the brain oscillations on SEEG data having both high temporal
(milliseconds) and spatial (millimeters) resolution.
The SEEG data were recorded from the ACC and the dorsolateral
PFC in drug-resistant epileptic patients engaged
in a standard visual oddball task. It was shown that compared
with the frequent stimuli, the rare (target) stimuli
and relative responses (stimulus counting and movement)
induced an increase of the DTF values from the ACC to
the dorsolateral PFC, which was statistically significant at
y rhythms (P < 0.05).
The first few questions raised by our results concern the
physiological meaning of y rhythms in the experimental
conditions present. It has been shown that event-related
modulation of y rhythms is an important neural correlate of
brain processes for the integration of sensorimotor and cognitive
information [Basar et al., 1999; Brankack et al., 1996;
Klimesch, 1999; Pfurtscheller and Lopes da Silva, 1999]. In
brief, y rhythms functionally connect the activity of hippocampal
systems and cortical mantle in many species [Buzsaki
and Draguhn, 2004], and are supposed to be related to
focused attention, working memory, encoding processes of
episodic memory, and control of action [Gevins and Smith,
2000; Klimesch, 1999; Klimesch et al., 2001, 2006; Sauseng
et al., 2002, 2004; Sochu˚rkova´ et al., 2006]. Furthermore, hippocampal
y rhythms are coordinated with several motor
patterns, including locomotion, orienting, rearing, and exploratory
whisking [Berg and Kleinfeld, 2003; Hasselmo
et al., 2002; Kleinfeld et al., 1999; Vanderwolf, 1969;
Whishaw and Schallert, 1977]. In light of these data, it can
be speculated that the flux of information within y rhythms
from the ACC (BAs 32, 24) to the dorsolateral PFC (BAs 9,
44, 45, 46) would be related to the attentional-control functions,
updating of working memory, and the control of
motor response to unpredictable rare (target) stimuli.
Indeed, here the mere observation of the frequent stimuli
was not able to modulate the directionality of the SEEG information
flux between the ACC and the dorsolateral PFC.
A second important issue raised by these results is the
physiological meaning of the unveiled directional information
flux from the ACC to the dorsolateral PFC. This
represents a step forward with respect to the notion of
‘‘coactivation/strict cooperation’’ between these areas previously
revealed by fMRI during oddball tasks [Bra´zdil
et al., 2007]. Indeed, such a directional information flux
supports the hypothesis of a peculiar triggering role of
the ACC over the dorsolateral PFC within cognitive functions.
When activated by sudden changes in the subject’s
environment, ACC may operate as a ‘‘supervisory attentional
system,’’ able to promote the activity of the dorsolateral
PFC for the integration of plans, instructions, stimulus
processing, decision making, short-term memory
updating, and response control [Shallice, 1988]. In this
sense, the present results extend previous evidence suggesting
that the directed attention is organized at the
level of a distributed large-scale network revolving
around three cortical epicentres (or local networks)—
ACC, dorsolateral PFC, and posterior parietal cortex
[Mesulam, 1981, 1990, 1999; Nebel et al., 2005], which
would provide a slightly different but interactive and
complementary type of top-down attentional functions
[Corbetta and Shulman, 2002; Fan et al., 2005; Kondo
et al., 2004; Konrad et al., 2005; Luks et al., 2002; MacDonald
et al., 2000; Milham et al., 2003].
The oddball task used in the present study can be
favorably employed in research of the attentional functions.
Rare target stimuli are well known to activate a
large network of regions as revealed by studies using
SEEG [Bra´zdil et al., 1999; Baudena et al., 1995; Halgren
et al., 1998] and fMRI (Bra´zdil et al., 2005; Clark et al.,
2000; Kiehl et al., 2001, 2005; Stevens et al., 2000; Yoshiura
et al., 1999]. Specifically, P3a potential of oddball tasks is
unambiguously generated in frontoparietocingulate system,
namely in dorsolateral PFC, supramarginal gyrus
and cingulate gyrus [Halgren et al., 1998]. All these
regions are significantly activated after targets in fMRI
studies too. In addition other cortical regions play important
role in target detection, including ventrolateral PFC,
medial temporal regions, intraparietal, and superior temporal
sulci as revealed consonantly by electrophysiological
and haemodynamic studies. All these regions are
involved in the complete set of cognitive processes resulting
in target detection and related execution. Their
involvement is nevertheless time-dependent to stimuli.
The P3 potential has a significantly shorter latency in
frontal sites than in parietal or temporal sites (but not in
ACC than in PFC). These literature findings suggest that
cognitive activation after targets is widespread and highly
organized in functional systems. In this line, the present
results should be interpreted as a demonstration of directionality
in the information processing flow within the
frontal lobe during oddball conditions, within a spatially
complex network of several areas. Indeed, correlations
and causalities could also be mediated by regions not
included in the present analysis. As a novel result of the
study, the results showed a central role of ACC in the
attentional functions and its direct functional connectivity
with PFC, in agreement with the strong evidence that the
dorsolateral prefrontal and cingulate cortices are highly
interconnected anatomically, and lesions in these sites are
associated with neglect and other attentional deficits
[Mesulam, 1990, 1999]. The extent of structural links
between ACC and lateral PFC is one of the most striking
features of cortico–cortical connectivity in the primate
frontal cortex. As pointed out by Barbas and Pandya
[1989], intrinsic connections of the cingulate cortices with
other frontocortical areas are not limited to immediate
neighbors, but also reach the more distant prefrontal
regions, particularly those in the dorsolateral PFC. It
seems therefore reasonable that in the task used, there is
an influence of ACC exerting over the PFC directly
through these anatomical interconnections. These data
importantly extend our previous fMRI analysis in which
directionality could be not determined at the level of cerebral
rhythms [Bra´zdil et al., 2007].
r Bra´zdil et al. r
r 144 r
CONCLUSIONS
Results of the present SEEG study showed that compared
with the frequent stimuli, the rare (target) stimuli of
an oddball task induced a frequency-specific increase of
directional information fluxes in humans (as revealed by
DTF of cerebral rhythms) from the ACC to the dorsolateral
PFC. These results suggest that when people have to react
to certain infrequent changes in the environment, the ACC
may affect the oscillatory activity of the dorsolateral PFC
by a selective frequency code. Future studies should define
the features of the stimuli able to modulate the mentioned
directional information fluxes in the two hemispheres.
The present results have an important heuristic values
and constitutes a basis for further experiments aimed at
addressing corollary working hypotheses. Does ACC interact
with dorsolateral PFC when subjects have to select
proper processes and response as a reaction to unpredictable
changes in the environment and/or when they simply
attend and perceive specific, albeit infrequent, relevant
objects? Are the relationships between ACC and dorsolateral
PFC affected by functional hemispherical asymmetry
or do they depend on the different specialization of area
24 or area 32 [BA 24 but not BA 32 is close to the projection
to motor areas; Bush et al., 2000]? Future experiments
should require subjects to reconfigure their response based
on changes in the environment (e.g., task switching) or
action value, the type of tasks commonly associated with
medial PFC activity. Here, we performed the first step
towards a challenging scientific venture.
REFERENCES
Babiloni C, Babiloni F, Carducci F, Cincotti F, Vecchio F, Cola B,
Rossi S, Miniussi C, Rossini PM (2004): Functional frontoparietal
connectivity during short-term memory as revealed by
high-resolution EEG coherence analysis. Behav Neurosci
118:687–697.
Babiloni C, Vecchio F, Cappa S, Pasqualetti P, Rossi S, Miniussi C,
Rossini PM (2006): Functional frontoparietal connectivity during
encoding and retrieval processes follows HERA model. A
high-resolution study. Brain Res Bull 68:203–212.
Banich MT, Milham MP, Atchley RA, Cohen NJ, Webb A, Wszalek
T, Kramer AF, Liang Z, Barad V, Gullet D, Shah C, Brown
C (2000): Prefrontal regions play a predominant role in imposing
an attentional ‘‘set’’: Evidence from fMRI. Cogn Brain Res 10:1–9.
Barbas H, Pandya DN (1989): Architecture and intrinsic connections
of the prefrontal cortex in the rhesus monkey. J Comp
Neurol 286:353–375.
Barch DM, Braver TS, Akbudak E, Conturo T, Ollinger J, Snyder
A (2001): Anterior cingulate cortex and response conflict: Effects
of response modality and processing domain. Cereb Cortex
11:837–848.
Basar E, Basar-Eroglu C, Karakas S, Schurmann M (1999): Are cognitive
processes manifested in event-related gamma, a, y and d
oscillations in the EEG? Neurosci Lett 259:165–168.
Baudena P, Halgren E, Heit G, Clarke JM (1995): Intracerebral
potentials to rare target and distractor auditory and visual
stimuli. III. Frontal cortex. Electroencephalogr Clin Neurophysiol
94:251–264.
Berg RW, Kleinfeld D (2003): Vibrissa movement elicited by rhythmic
electrical microstimulation to motor cortex in the aroused
rat mimics exploratory whisking. J Neurophysiol 90:2950–2963.
Blinowska KJ, Kus R, Kaminski M. (2004): Granger causality and
information flow in multivariate processes. Phys Rev E Stat
Nonlin Soft Matter Phys 70(5, Part 1):050902.
Botvinick M, Nystrom LE, Fissell K, Carter CS, Cohen JD (1999):
Conflict monitoring versus selection-for-action in anterior cingulate
cortex. Nature 402:179–181.
Brankack J, Seidenbecher T, Muller-Gartner HW (1996): Task-relevant
late positive component in rats: is it related to hippocampal
y rhythm? Hippocampus 6:475–482.
Braver TS, Barch DM, Kelley WM, Buckner RL, Cohen NJ, Miezin
FM, Snyder AZ, Ollinger JM, Akbudak E, Conturo TE, Petersen
SE (2001): Direct comparison of prefrontal cortex regions
engaged by working and long-term memory tasks. Neuroimage
14:48–59.
Bra´zdil M, Rektor I, Dufek M, Daniel P, Jura´k P, Kuba R (1999):
The role of frontal and temporal lobes in visual discrimination
task—Depth ERP studies. Neurophysiol Clin 29:339–350.
Bra´zdil M, Dobsˇı´k M, Mikl M, Hlusˇtı´k P, Daniel P, Pazˇourkova´ M,
Krupa P, Rektor I (2005): Combined event-related fMRI and intracerebral
ERP study of an auditory oddball task. Neuroimage
26/1:285–293.
Bra´zdil M, Mikl M, Marecˇek R, Krupa P, Rektor I. (2007): Effective
connectivity in target stimulus processing: A dynamic causal
modeling study of visual oddball task. NeuroImage 35:827–835.
Bush G, Luu P, Posner MI (2000): Cognitive and emotional influences
in anterior cingulate cortex. Trends Cogn Sci 4:215–222.
Buzsaki G, Draguhn A (2004): Neuronal oscillations in cortical networks.
Science 304:1926–1929.
Carter CS, Braver TS, Barch DM, Botvinick MM, Noll D, Cohen
JD (1998): Anterior cingulate cortex, error detection, and the
online monitoring of performance. Science 280:747–749.
Clark VP, Fannon S, Lai S, Benson R, Bauer L (2000): Responses to
rare visual target and distractor stimuli using event-related
fMRI. J Neurophysiol 83:3133–3139.
Cohen JD, Botvinick M, Carter CS (2000): Anterior cingulate and
prefrontal cortex: Who’s in control? Nat Neurosci 3:421–423.
Corbetta M, Shulman GL (2002): Control of goal-directed and stimulus-driven
attention in the brain. Nat Rev Neurosci 3:201–215.
De Gennaro L, Vecchio F, Ferrara M, Curcio G, Rossini PM, Babiloni
C (2004): Changes in fronto-posterior functional coupling
at sleep onset in humans. J Sleep Res 13:209–217.
Fan J, McCandliss BD, Fossella J, Flombaum JI, Posner MI (2005):
The activation of attentional networks. Neuroimage 26:471–479.
Gevins A, Smith ME (2000): Neurophysiological measures of
working memory and individual differences in cognitive ability
and cognitive style. Cereb Cortex 10:829–839.
Gitelman DR, Nobre AC, Parrish TB, LaBar KS, Kim YH, Meyer
JR, Mesulam MM (1999): A large-scale distributed network for
covert spatial attention: Further anatomical delineation based
on stringent behavioral and cognitive controls. Brain 122:1093–
1106.
Halgren E, Marinkovic K, Chauvel P (1998): Generators of the late
cognitive potentials in auditory and visual oddball tasks. Electroencephalogr
Clin Neurophysiol 106:156–164.
Hasselmo ME, Hay J, Ilyn M, Gorchetchnikov A (2002): Neuromodulation,
y rhythm and rat spatial navigation. Neural Netw
15:689–707.
Kaminski MJ, Blinowska KJ (1991): A new method of the description
of the information flow in the structures. Biol Cybern
65:203–210.
r Effective Connectivity in Rare Stimuli Processing r
r 145 r
Kaminski MJ, Blinowska KJ, Szclenberger W (1997): Topographic
analysis of coherence and propagation of EEG activity during
sleep and wakefulness. Electroencephalogr Clin Neurophysiol
102:216–227.
Kiehl KA, Laurens KR, Duty TL, Forster BB, Liddle PF (2001):
Neural sources involved in auditory target detection and novelty
processing: An event-related fMRI study. Psychophysiology
38:133–142.
Kiehl KA, Stevens MC, Laurens KR, Pearlson G, Calhoun VD, Liddle
PF (2005): An adaptive reflexive processing model of neurocognitive
function: Supporting evidence from a large scale (n
5 100) fMRI study of an auditory oddball task. Neuroimage
25:899–915.
Kleinfeld D, Berg RW, O’Connor SM (1999): Anatomical loops and
their electrical dynamics in relation to whisking by rat.
Somatosens Mot Res 16:69–88.
Klimesch W (1999): EEG a and y oscillations reflect cognitive and
memory performance: A review and analysis. Brain Res Brain
Res Rev 29:169–195.
Klimesch W, Doppelmayr M, Yonelinas A, Kroll NE, Lazzara M,
Rohm D, Gruber W (2001): y Synchronization during episodic
retrieval: Neural correlates of conscious awareness. Brain Res
Cogn Brain Res 12:33–38.
Klimesch W, Doppelmayr M, Hanslmayr S (2006): Upper a ERD
and absolute power: Their meaning for memory performance.
Prog Brain Res 159:151–165.
Kondo H, Osaka N, Osaka M (2004): Cooperation of the anterior
cingulate cortex and dorsolateral prefrontal cortex for attention
shifting. Neuroimage 23:670–679.
Konrad K, Neufang S, Thiel CM, Specht K, Hanisch C, Fan J, Herpertz-Dahlmann
B, Fink GR (2005): Development of attentional
networks: An fMRI study with children and adults. Neuroimage
28:429–439.
Luks TL, Simpson GV, Feiwell RJ, Miller WL (2002): Evidence for
anterior cingulate cortec involvement in monitoring preparatory
attentional set. NeuroImage 17:792–802.
MacDonald AW, Cohen JD, Stenger VA, Carter CS (2000):
Dissociating the role of the dorsolateral prefrontal and anterior
cingulate cortex in cognitive control. Science 288:1835–
1838.
Mesulam MM (1981): A cortical network for directed attention
and unilateral neglect. Ann Neurol 10:309–325.
Mesulam MM (1990): Large-scale neurocognitive networks and
distributed processing for attention, language, and memory.
Ann Neurol 28:597–613.
Mesulam MM (1999): Spatial attention and neglect: parietal, frontal
and cingulate contributions to the mental representation
and attentional targeting of salient extrapersonal events. Philos
Trans R Soc London 354:1325–1346.
Milham MP, Banich MT, Claus ED, Cohen NJ (2003): Practicerelated
effects demonstrate complementary roles of anterior
cingulate and prefrontal cortices in attentional control. Neuroimage
18:483–493.
Nebel K, Wiese H, Stude P, de Greiff A, Diener HC, Keidel
M (2005): On the neural basis of focused and divided attention.
Cogn Brain Res 25:760–776.
Pfurtscheller G, Lopes da Silva FH (1999): Event-related EEG/
MEG synchronization and desynchronization: Basic principles.
Clin Neurophysiol 110:1842–1857.
Sauseng P, Klimesch W, Gruber WR, Doppelmayr M, Stadler M,
Schabus M (2002): The interplay between y and a oscillations in
the human electroencephalogram reflects the transfer of information
between memory systems. Neurosci Lett 323:121–124.
Sauseng P, Klimesch W, Doppelmayr M, Hanslmayr S, Schabus
M, Gruber WR (2004): y Coupling in the human electroencephalogram
during a working memory task. Neurosci Lett
354:123–126.
Shallice T (1988): From Neuropsychology to Mental Structure.
Cambridge: Cambridge University Press.
Shallice T, Burgess PW, Schon F, Baxter DM (1989): The origin of
utilization behavior. Brain 112:1587–1598.
Snyder E, Hillyard SA (1976): Long-latency evoked potentials to
irrelevant, deviant stimuli. Behav Biol 16:319–331.
Sochu˚rkova´ D, Bra´zdil M, Jura´k P, Rektor I (2006): P3 and ERD/
ERS in a visual oddball paradigm. A depth EEG study from
the mesial temporal structures. J Psychophysiol 20:32–39.
Squires NK, Squires KC, Hillyard SA (1975): Two varieties of
long-latency positive waves evoked by unpredictable auditory
stimuli in man. Electroencephalogr Clin Neurophysiol 38:387–
401.
Stevens AA, Skudlarski P, Gatenby JC, Gore JC (2000): Eventrelated
fMRI of auditory and visual oddball tasks. Magn Reson
Imaging 18:495–502.
Talairach J, Szikla G, Tournoux P, Prosalentis A, Bordas-Ferrer M,
Covello J (1967): Atlas d’anatomie stereotactique du telencephale.
Paris: Masson
Vanderwolf CH (1969): Hippocampal electrical activity and voluntary
movement in the rat. Electroencephalogr Clin Neurophysiol
26:407–418.
Verleger R, Jaskowski P, Wascher E (2005): Evidence for an integrative
role of P3b in linking reaction to perception. J Psychophysiol
19:165–181.
Whishaw IQ, Schallert T (1977): Hippocampal RSA (y), apnea, bradycardia
and effects of atropine during underwater swimming
in the rat. Electroencephalogr Clin Neurophysiol 42:389–396.
Woldorf MG, Fichtenholtz HM, Song AW, Mangun GR. (2001):
Cue- and target-related processing in a fast-rate cued visual
spatial attention paradigm. Presented at the 2001 Meeting of
the Organization for Human Brain Mapping. Brighton, UK.
Yoshiura T, Zhong J, Shibata DK, Kwok WE, Shrier DA, Numaguchi
Y (1999): Functional MRI study of auditory and visual oddball
tasks. Neuroreport 10:1683–1688.
r Bra´zdil et al. r
r 146 r
124
V. Seznam citované literatury
Allison T, McCarthy G, Wood CC, Darcey TM, Spencer DD, Williamson PD. Human cortical
potentials evoked by stimulation of the median nerve. II. Cytoarchitectonic areas
generating short-latency activity. J Neurophysiol 1989;62:694-710.
Allison T, Puce A, Spencer DD, McCarthy G. Electrophysiological studies of human face
perception. I: Potentials generated in occipitotemporal cortex by face and non-face stimuli.
Cereb Cortex 1999;9:415-430.
Allman JM, Hakeem A, Erwin JM, Nimchinsky E, Hof P. The anterior cingulate cortex. The
evolution of an interface between emotion and cognition. Ann NY Acad Sci 2001;935:107-
117.
Amador N, Fried I. Single-neuron activity in the human supplementary motor area underlying
preparationor action. J Neurosurg 2004;100:250-259.
Amaral DG, Lavenex P. Hippocampal neuroanatomy. In: Andersen P, Morris R, Amaral DG,
Bliss T, O’Keefe J (eds.). The Hippocampus Book. Oxford: Oxford University Press.
2007;pp. 37–114.
Anllo-Vento L, Luck SJ, Hillyard SA. Spatio-temporal dynamics of attention to color: Evidence
from human electrophysiology. Hum Brain Mapp 1998;6(4):216–38.
Antal A, Kéri S, Kovács G, Janka Z, Benedek G. Early and late components of visual
categorization: an event-related potential study. Cogn Brain Res 2000;9:117-119.
Axmacher N, Cohen MX, Fell J, Haupt S, Dümpelmann M, Elger CE, Schlaepfer TE, Lenartz
D, Sturm V, Ranganath CH. Intracranial EEG correlates of expectancy and memory
formation in the human hippocampus and nucleus accumbens. Neuron 2010; 65(4):541-
549.
Baas JM, Kenemans JL, Mangun GR. Selective attention to spatial frequency: an ERP and
source localization analysis. Clin Neurophysiol 2002;113(11):1840–1854.
Babiloni C, Vecchio F, Cappa S, Pasqualetti P, Rossi S, Miniussi C, Rossini PM. Functional
frontoparietal connectivity during encoding and retrieval processes follows HERA model.
A high-resolution study. Brain Res Bull 2006;68:203-212.
Baccala L, Sameshima K. Partial Directed Coherence: A new conception in neural structure
determination. Biol Cybern 2001;84(6);463-74.
Bancaud J. Apport de l´exploration fonctionelle par voie stéréotaxique á la chirurgie de
l´épilepsie. Neurochirurgie 1959;5:55-112.
Barch DM, Braver TS, Akbudak E, Conturo T, Ollinger J, Snyder A. Anterior cingulate cortex
and response conflict: Effects of response modality and processing domain. Cereb Cortex
2001;11:837-848.
Başar E. Brain functions and oscillations. Vol. 1. Brain Oscillations – Principles and
Approaches. Springer, Berlin, Heidelberg. 1998.
Başar E, Başar-Eroglu C, Demiralp T, Ademoglu A, Schürmann M. Oscillatory brain dynamics,
wavelets and cognition. Brain Lang 1999a;66:146-183.
125
Basar E, Basar-Eroglu C, Karakas S, Schurmann M. Are cognitive processes manifested in
event-related gamma, α, θ and δ oscillations in the EEG? Neurosci Lett 1999b;259:165-
168.
Bates AT, Liddle PF, Kiehl KA, Ngan ETC. State dependent changes in error monitoring in
schizophrenia. J Psychiatr Res 2004; 38(3):347-356.
Baudena P, Halgren E, Heit G, Clarke JM. Intracerebral potentials to rare target and distractor
auditory and visual stimuli. III. Frontal cortex. Electroenceph Clin Neurophysiol
1995;94:251-264.
Bechara A, Damasio H, Damasio AR. Emotion, decision making and the orbitofrontal cortex.
Cereb Cortex 2000;10:295-307.
Berger H. Ueber das Elektroenkephalogramm des Menschen. Arch Psychiatr Nervenkr
1929;87:527-570.
Bland BH, Oddie SD. Theta band oscillation and synchrony in the hippocampal formation and
associated structures: The case for its role in sensorimotor integration. Behav Brain Res
2001;127(1-2):119-136.
Botvinick MM, Braver TS, Carter CS, Barch DM, Cohen JD. Evaluating the demand for
control: Anterior cingulate cortex and crosstalk monitoring. Psychol Rev 2001;108:624-
652.
Boutros NN, Mears R, Pflieger ME, Moxon KA, Ludowig E, Rosburg T. Sensory gating in the
human hippocampal and rhinal regions: Regional differences. Hippocampus
2008;18(3):310-316.
Brandt ME. Visual and auditory evoked phase resetting of the alpha EEG. Int J Psychophysiol
1997;26(1-3):285-98.
Brankack J, Seidenbecher T, Muller-Gartner HW. Task-relevant late positive component in
rats: is it related to hippocampaly rhythm? Hippocampus 1996;6:475-482.
Braver TS, Barch DM, Gray JR, Molfese DL, Snyder A. Anterior cingulate cortex and response
conflict: Effects of frequency, inhibition, and errors. Cereb Cortex 2001;11: 825-836.
Brázdil M, Rektor I, Dufek M, Daniel P, Jurák P, Kuba R. The role of frontal and temporal
lobes in visual discrimination task - depth ERP studies. Neurophysiol Clin 1999;29:339-
350.
Brázdil M, Rektor I, Daniel P, Dufek M, Jurák P. Intracerebral event-related potentials to
subthreshold target stimuli. Clin Neurophysiol 2001;112(4):650-661.
Brázdil M, Roman R, Falkenstein M, Daniel P, Jurák P, Rektor I. Error processing – evidence
from intracerebral ERP recordings. Exp Brain Res 2002;146:460-466.
Brázdil M, Roman R, Daniel P, Rektor I. Intracerebral somatosensory event-related potentials:
Effect of response type (button pressing versus mental counting) on P3-like potentials
within the human brain. Clin Neurophysiol 2003;114(8):1489-1496.
Brázdil M, Mikl M, Mareček R, Krupa P, Rektor I. Effective connectivity in target stimulus
processing: A dynamic causal modeling study of visual oddball task. NeuroImage
2007;35:827-835.
Bressler SL and Kelso JA. Cortical coordination dynamics and cognition. Trends Cogn Sci
2001;5:26-36.
126
Brovelli A, Lachaux JP, Kahane P, Boussaoud D. High gamma frequency oscillatory activity
dissociates attention from intention in the human premotor cortex. NeuroImage
2005;28:154-164.
Botvinick M, Nystrom LE, Fissell K, Carter CS, Cohen JD. Conflict monitoring versus
selection-for-action in anterior cingulate cortex. Nature 1999;402:179-181.
Bush G, Luu P, Posner MI. Cognitive and emotional influences in anterior cingulate cortex.
Trends Cogn Sci 2000;4:215-222.
Carter CS, Braver TS, Barch DM, Botvinick MM, Noll D, Cohen JD. Anterior cingulate cortex,
error detection, and the online monitoring of performance. Science 1998;280(5364):747-
749.
Clark VP, Fannon S, Lai S, Benson R, Bauer L. Responses to rare visual target and distractor
stimuli using event-related fMRI. J Neurophysiol 2000;83:3133-3139.
Cohen JD, Botvinick M, Carter CS. Anterior cingulate and prefrontal cortex: Who’s in control?
Nat Neurosci 2000;3:421-423.
Coles M, Scheffers M, Holroyd C. Why is there an ERN/Ne on correct trials? Response
representations, stimulus-related components, and the theory of error-processing. Biol
Psychol 2001;56(3):173-189.
Comerchero MD, Polich J. P3a, perceptual distinctiveness, and stimulus modality. Cogn Brain
Res 1998;7:41-48.
Cracco RQ, Cracco JB. Somatosensory evoked potentials in man: Far-field potentials.
Electroenceph Clin Neurophysiol 1976;41:460-466.
Crone NE, Miglioretti DL, Gordon B, Lesser RP. Functional mapping of human sensorimotor
cortex with electrocorticographic spectral analysis. II. Event-related synchronization in the
gamma band. Brain 1998a;121(12):2301-2315.
Crone NE, Miglioretti DL, Gordon B, Sieracki JM, Wilson MT, Uematsu S, Lesser R.
Functional mapping of human sensorimotor cortex with electrocorticographic spectral
analysis. I. Alpha and beta event-related desynchronization. Brain 1998b; 121(12):2271-
2299.
Crone NE, Boatman D, Gordon B, Hao L. Induced electrocorticographic gamma activity during
auditory perception. Clin Neurophysiol 2001;112(4):565-582.
Desmurget M, Reilly KT, Richard N, Szathmari A, Mottolese C, Sirigu A. Movement intention
after parietal cortex stimulation in humans. Science 2009;324:811-813.
Devinsky O, Morrel MJ, Vogt BA. Contributions of anterior cingulate cortex to behavior. Brain
1995;118:279-306.
Donchin E, Ritter W, McCallum WC. Cognitive psychophysiology: The endogenous
components of the ERP. In: Callaway E, Tueting P, Koslow SH (eds.). Event-Related Brain
Potentials in Man. Academic Press, New York, 1978;pp. 349-411.
Doucet C, Stelmack RM. The effect of response execution on P3 latency, reaction time, and
movement time. Psychophysiol 1999;36:351-363.
Eichenbaum H, Sauvage M, Fortin N, Komorowski R, Lipton P. Towards a functional
organization of episodic memory in the medial temporal lobe. Neurosci Biobehav Rev
2012;36(7):1597-1608.
127
Ekstrom AD, Kahana MJ, Caplan JB, Fields TA, Isham EA. Cellular networks underlying
human spatial navigation. Nature 2003;425:184-188.
Ekstrom AD, Caplan JB, Ho E, Shattuck K, Fried I, Kahana MJ. Human hippocampal theta
activity during virtual navigation. Hippocampus 2005;15(7):881-889.
Falkenstein M, Hohnsbein J, Hoormann J. Effects of crossmodal divided attention on late ERP
components. II. Error processing in choice reaction tasks. Electroenceph Clin Neurophysiol
1991;78:447-455.
Falkenstein M, Hoormann J, Christ S, Hohnsbein J. ERP components on reaction errors and
their functional significance: A tutorial. Biol Psychol 2000;51:87-107.
Fiehler K, Ullsperger M, von Cramon DY. Neural correlates of error detection and error
correction: Is there a common neuroanatomical substrate? Eur J Neurosci 2004;19:3081-
3087.
Fell J, Dietl T, Grunwald T, Kurthen M, Klaver P, Trautner P Schaller C, Elger CE, Fernández
G. Neural bases of cognitive ERPs: More than phase reset. J Cogn Neurosci
2004;16(9):1595-1604.
Fell J, Köhling R, Grunwald T, Klaver P, Dietl T, Schaller C, Becker A, Elger CE, Fernández
G. Phase-locking characteristics of limbic P3 responses in hippocampal sclerosis.
Neuroimage 2005;24(4):980-989.
Fernández G, Klaver P, Fell J, Grunwald T, Elger CE. Human declarative memory formation:
Segregating rhinal and hippocampal contributions. Hippocampus 2002;12(4):514-519.
Fried I, Katz A, McCarthy G, Sass KJ, Williamson P, Spencer SS, Spencer DD. Functional
organization of human supplementary motor cortex studied by electrical stimulation. J
Neurosci 1991;11:3656-3666.
Fried I, Wilson CL, Maidment NT, Engel J, Behnke E, Fields TA, Macdonald KA, Morrow
JW, Ackerson L. Cerebral microdialysis combined with single-neuron and
electroencephalographic recording in neurosurgical patients. Technical note. J Neurosurg
1999; 91:697-705.
Fries P, Reynolds J H, Rorie A E, Desimone R. Modulation of oscillatory neuronal
synchronization by selective visual attention. Science 2001;291:1560-1563.
Garavan H, Ross TJ, Kaufman J, Stein EA. A midline dissociation between error-processing
and response conflict monitoring. NeuroImage 2003;20:1132-1139.
Gehring WJ, Gratton G, Coles MGH, Donchin E. Probability effects on stimulus evaluation
and response processes. J Exp Psychol Hum Percept Perform 1992;18:198-216.
Gehring WJ, Goss B, Coles MGH, Meyer DE, Donchin E. A neural system for error detection
and compensation. Psychological Science 1993;4:385-390.
Gehring WJ, Knight, RT. Prefrontal-cingulate interactions in action monitoring. Nat Neurosci
2000;3:516-520.
Granger C. Investigating causal relations by econometric models and cross-spectral methods.
Econometrica, 1969, vol. 37, pp. 424-438.
Greenlee JD, Jackson AW, Chen F, Larson CR, Oya H, Kawasaki H, Chen H, Howard MA,
Gribble PL. Human auditory cortical activation during self-vocalization. PLoS One
2011;6:e14744.
128
Gunter TC, Wijers AA, Jackson JL, Mulder G. Visual spatial attention to stimuli presented on
the vertical and horizontal median: An ERP study. Psychophysiol 1994;31:140-153.
Gurtubay I, Alegre M, Labarga A, Malanda A, Iriarte J, Artieda J. Gamma band activity in an
auditory oddball paradigm studied with the wavelet transform. Clin Neurophysiol
2001;112(7):1219-28.
Grunwald T, Elger CE, Lehnertz K, Van Roost D, Heinze HJ. Alterations of intrahippocampal
cognitive potentials in temporal lobe epilepsy. Electroenceph Clin Neurophysiol
1995;95(1):53-62.
Grunwald T, Lehnertz K, Heinze HJ, Helmstaedter C, Elger CE. Verbal novelty detection
within the human hippocampus proper. Proc Natl Acad Sci USA 1998;95(6):3193-3197.
Hajcak G, Moser JS, Yeung N, Simons RF. On the ERN and the significance of errors.
Psychophysiology 2005;42(2):151-160.
Halgren E, Baudena P, Clarke JM, Heit G, Liégeois C, Chauvel P, Musolino A. Intracerebral
potentials to rare target and distractor auditory and visual stimuli. I. Superior temporal
plane and parietal lobe. Electroenceph Clin Neurophysiol 1995a;94:191-220.
Halgren E, Baudena P, Clarke JM, Heit G, Marinkovic K, Deveaux B, Vignal J, Biraben A.
Intracerebral potentials to rare target and distractor auditory and visual stimuli. II. Medial,
lateral and posterior temporal lobe. Electroenceph Clin Neurophysiol 1995b;94:229-250.
Halgren E, Marinkovic K, Chauvel P. Generators of the late cognitive potentials in auditory
and visual oddball tasks, Electroenceph Clin Neurophysiol 1998;106:156-164.
Hillyard SA, Picton TW, Regan D. Sensation, perception and attention: Analysis using ERPs.
In: Callaway E, Tueting P, Koslow SH (eds.). Event-Related Brain Potentials in Man.
Academic Press, New York, 1978;pp. 223-321.
Hillyard SA, Vogel EK, Luck SJ. Sensory gain control (amplification) as a mechanism of
selective attention: electrophysiological and neuroimaging evidence. The Royal Society
1998a;353:1257-1270.
Hillyard SA, Anllo-Vento L. Event-related brain potentials in the study of visual selective
attention. Proc Natl Acad Sci 1998b;95:781-787.
Hochberg LR, Serruya MD, Friehs GM, Mukand JA, Saleh M, Caplan AH, Branner A, Chen
D, Penn RD, Donoghue JP. Neuronal ensemble control of prosthetic devices by a human
with tetraplegia. Nature 2006;442:164-171.
Holroyd CB, Dien J, Coles MGH. Error-related scalp potentials elicited by hand and foot
movements: Evidence for an output-independent error-processing system in humans.
Neurosci Lett 1998;242:65-68.
Howard MA, Volkov IO, Abbas PJ, Damasio H, Ollendieck MC, Granner MA. A chronic
microelectrode investigation of the tonotopic organization of human auditory cortex. Brain
Res 1996a;724:260-264.
Howard MA, Volkov IO, Granner MA, Damasio HM, Ollendieck MC, Bakken HE. A hybrid
clinical-research depth electrode for acute and chronic in vivo microelectrode recording of
human brain neurons. Technical note. J. Neurosurg 1996b;84:129-132.
Ishikawa O. Electroencephalographical study of human thalamus. Folia Psychiatr Neurol Jpn
1957;11:128-149.
129
Jacobs J, Kahana MJ, Ekstrom AD, Mollison MV, Fried I. A sense of direction in human
entorhinal cortex. Proc Natl Acad Sci USA 2010;107:6487-6492.
Jasper H, Carmichael L. Electrical potentials from the intact human brain. Science 1935;81: 51.
Jensen O, Kaiser J, Lachaux J P. Human gamma-frequency oscillations. associated with
attention and memory. Trends Neurosci 2007;30,317-324.
Jentzsch I, Sommer W. Sequence-sensitive subcomponents of P300: topographical analyses
and dipole source localization. Psychophysiol 2001;38(4):607-621.
Jewett DL, Romano MN, Williston JS. Human auditory evoked potentials: Possible brainstem
components detected on the scalp. Science 1970;167:1517-1518.
Klimesch W. Memory processes, brain oscillations and EEG synchronization. J Psychophysiol
1996;24:61-100.
Klimesch W. EEG alpha and theta oscillations reflect cognitive and memory performance: A
review and analysis. Brain Res Rev 1999;29:169-195.
Klimesch W, Doppelmayr M, Rhöm, D, Pöllhuber D, Stadler W. Simultaneous
desynchronization and synchronization of different alpha responses in the human
electroencephalograph: a neglected paradox? Neurosci Lett 2000;284:97-100.
Knight RT, Scabini D. Anatomic bases of event-related potentials and their relationship to
novelty detection in humans. Clin Neurophysiol 1998;15(1):3-13.
Knott JR, Hayne RA, Meyers HR. Physiology of sleep wave characteristics and temporal
relations of human electroencephalograms recorded from the thalamus, the corpus striatum
and the surface of the scalp. Arch Neurol Psychiatry (Chicago) 1950;63:526-527.
Kreiman G, Koch C, Fried I. Category-specific visual responses of single neurons in the human
medial temporal lobe. Nat Neurosci 2000a;3:946-953.
Kreiman G, Koch C, Fried I. Imagery neurons in the human brain. Nature 2000b;408:357-361.
Krolak-Salmon P, Henaff MA, Isnard J, Tallon-Baudry C, Guenot M, Vighetto A, Bertrand O,
Mauguiere F. An attention modulated response to disgust in human ventral anterior insula.
Ann Neurol 2003;53:446-453.
Krolak-Salmon P, Henaff MA, Vighetto A, Bertrand O, Mauguiere F. Early amygdala
associated with attention and memory. Trends Neurosci 2004;30:317-324.
Kukleta M, Brazdil M, Roman R, Bob P, Rektor I. Cognitive network interactions and beta 2
coherence in processing non-target stimuli in visual oddball task. Physiol Res 2009;58:139-
148.
Lachaux JP, Rudrauf D, Kahane P. Intracranial EEG and human brain mapping. Journal of
Physiology-Paris 2003;97(4-6):613-28.
Lachaux JP, Fonlupt P, Kahane P, Minotti L, Hoffmann D. Relationship between task-related
gamma oscillations and BOLD signal: new insights from combined fMRI and intracranial
EEG. Hum Brain Mapp 2007;28:1368-1375.
Lachaux J-P, Axmacher N, Mormann F, Halgren E, Crone NE. High-frequency neural activity
and human cognition: Past, present and possible future of intracranial EEG research. Prog
Neurobiol 2012;98(3):279-301.
130
Luck SJ. Measuring ERP latencies. In: Luck SJ, editor. An introduction to the event-related
potential technique. Cambridge: MIT Press 2005;pp. 237-249.
Ludowig E, Bien CG, Elger CE, Rosburg T. Two P300 generators in the hippocampal
formation. Hippocampus 2010;20(1):186-195.
Luu P, Flaisch T, Tucker DM. Medial frontal cortex in action monitoring. J Neurosci
2000;20:464-469.
Martínez A, DiRusso F, Anllo-Vento L, Hillyard SA. Electrophysiological analysis of cortical
mechanisms of selective attention to high and low spatial frequencies. Clin Neurophysiol
2001a;112:1980-1998.
Martínez A, DiRusso F, Anllo-Vento L, Sereno MI, Buxton RB, Hillyard SA. Putting spatial
attention on the map: timing and localization of stimulus selection processes in striate and
extrastriate visual areas. Vision Res 2001b;41:1437-1457.
Marusič P, Kuba R, Vojtěch Z. Elektroencefalografie. In: Brázdil M, Marusič P (eds.). Epilepsie
temporálního laloku. Praha: Triton.2006;pp. 122-155.
Mathalon DH, Fedor M, Faustman WO, Gray M, Askari N, Ford JM. Response-monitoring
dysfunction in schizophrenia: An event-related brain potential study. J Abnorm Psychol
2002;111(1): 22-41.
Mazaheri A, Picton TW. EEG spectral dynamics during discrimination of auditory and visual
targets. Cogn Brain Res 2005;25:81-96.
McCarthy G, Wood CC, Williamson PD, Spencer DD. Task-dependent field potentials in
human hippocampal formation. J Neurosci 1989;9(12):4253-68.
McCarthy G, Luby M, Gore J, Goldman-Rakic P. Infrequent events transiently activate human
prefrontal and parietal cortex as measured by functional MRI. J Neurophysiol
1997;77:1630-1634.
Mecklinger A, Kramer AF, Strayer DL. Event-related potentials and EEG components in a
semantic memory search task. Psychophysiol 1992;29:104-119.
Meckler C, Allain S, Carbonnell L, Hasbroucq T, Burle B, Vidal F. Executive control and
response expectancy: A laplacian ERP study. Psychophysiology 2011;48(3):303-311.
Menon V, Adleman NE, White CD, Glover GH, Reiss AL. Error-related brain activation during
a Go/NoGo response inhibition task. Hum Brain Mapp 2001;12:131-143.
Mesulam MM. Large-scale neurocognitive networks and distributed processing for attention,
language, and memory. Ann Neurol 1990;28:597-613.
Meyers HR, Hayne R. Electrical potentials of the corpus striatum and cortex in Parkinsonism
and hemiballism. Trans Am Neurol Assoc 1948;73:10-14.
Molholm S, Ritter W, Murray MM, Javitt DC, Schroeder CE, Foxe JJ. Multisensory auditory –
visual interactions during early sensory processing in humans: a high-density electrical
mapping study. Cogn Brain Res 2002;14(1):115-28.
Mormann F, Lehnertz K, David P, Elger EC. Mean phase coherence as a measure for phase
synchronization and its application to the EEG of epilepsy patients. Physica D: Nonlinear
Phenomena, 2000, vol. 144, pp. 358–369.
131
Mukamel R, Ekstrom A, Kaplan J, Iacoboni M, Fried I. Single-neuron responses in humans
during execution and observation of actions. Curr Biol 2010a;20:1-7.
Mukamel R, Fried I. Human Intracranial Recordings and Cognitive Neuroscience. Annu Rev
Psychol 2012;63(1):511-537.
Niedermayer E. Depth electroencephalography. In: Niedermayer E, Lopes da Silva F (Eds).
Electroencephalography: Basic principles, clinical applications, and related fields. 5th ed.,
Philadelphia: Lippincott Williams & Wilkins. 2005;pp. 733-748.
Novak GP, Ritter W, Vaughan HG Jr, Wiznitzer ML. Differentiation of negative event-related
potentials in an auditory discriminatory task. Electroenceph Clin Neurophysiol 1990
75:255-275.
Ojemann G, Ojemann J, Lettich E, Berger M. Cortical language localization in left, dominant
hemisphere: an electrical stimulation mapping investigation in 117 patients. J Neurosurg
1989;71:316-326.
Ojemann GA, Schoenfield-McNeill J, Corina DP. Anatomic subdivisions in human temporal
cortical neuronal activity related to recent verbal memory, Nat Neurosci 2002;5:64-71.
Oya H, Kawasaki H, Howard MA, Adolphs R. Electrophysiological responses in the human
amygdala discriminate emotion categories of complex visual stimuli. J Neurosci
2002;22:9502-9512.
Paller KA, Kutas M, Mayes AR. Neural correlates of encoding in an incidental learning
paradigm. Electroenceph Clin Neurophysiol 1987;67:360-371.
Paller KA, McCarthy G. Field potentials in the human hippocampus during the encoding and
recognition of visual stimuli. Hippocampus 2002;12(3):415-420.
Paus T. Primate anterior cingulate cortex: Where motor control, drive and cognition interface.
Nat Rev Neurosci 2001;2:417-424.
Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man
as studied by electrical stimulation. Brain 1937;60:389-443.
Penfield W, Baldwin M. Temporal lobe seizures and the technic of subtotal temporal
lobectomy. Ann Surg 1952;136:625-634.
Penfield W, Roberts L. Speech and Brain Mechanisms. Princeton, NJ: Princeton Univ. Press.
1959.
Pfefferbaum A, Christensen C, Ford JM, Kopell BS. Apparent response incompatibility effects
on P300 latency depend on the task. Electroenceph Clin Neurophysiol 1986;64:424-437.
Pfurtscheller G. Event-related synchronization (ERS): an electrophysiological correlate of
cortical areas at rest. Electroenceph Clin Neurophysiol 1992;83;p. 6269.
Polich J. Attention, probability, and task demands as determinants of P300 latency from
auditory stimuli. Electroenceph Clin Neurophysiol 1986;63:251-259.
Polich J. P300 from a passive auditory paradigm. Electroenceph Clin Neurophysiol
1989;74(4):312-320.
Polich J. Updating P300: An integrative theory of P3a and P3b. Clin Neurophysiol
2007;118(10):2128-48.
132
Posner MI, Rothbart MK, Sheese BE, Tang Y. The anterior cingulate gyrus and the mechanism
of self-regulation. Cogn Affect Behav Ne 2007;7:391-395.
Privman E, Nir Y, Kramer U, Kipervasser S, Andelman F, Neufeld MY, Mukamel R, Yeshurun
Y, Fried I, Malach R. Enhanced category tuning revealed by intracranial
electroencephalograms in high-order human visual areas. J Neurosci 2007;27:6234-6242.
Proverbio AM, Esposito P, Zani A. Early involvement of the temporal area in attentional
selection of grating orientation: an ERP study. Cogn Brain Res 2002;13:139-151.
Rappelsberger P, Lacroix D, Petsche H. Amplitude and coherence mapping: Its application in
psycho- and pathophysiological studies. In: Rother M, Zwiener U (eds.). Quantitative EEG
analysis – clinical and new methods. Universitätsverlag Jena 1993;pp. 179-186.
Rasmussen T, Jasper, HH. Temporal lobe epilepsy: Indication for operation and surgical
technique. In Bailey P, Baldwin M (eds.). Temporal Lobe Epilepsy. Springfield, IL:
Charles C Thomas.1958; pp. 440-460.
Ridderinkhof KR, Ullsperger M, Crone EA, Nieuwenhuis S. The role of the medial frontal
cortex in cognitive control. Science 2004;306:443-447.
Roger C, Bénar CG, Vidal F, Hasbroucq T, Burle B. Rostral cingulate zone and correct response
monitoring: ICA and source localization evidences for the unicity of correct- and errornegativities.
NeuroImage 2010;51(1):391-403.
Roman R, Brázdil M, Jurák P, Rektor I, Kukleta M. Intracerebral P3-like waveforms and the
length of the stimulus-response interval in a visual oddball paradigm. Clin Neurophysiol
2005;116(1):160-171.
Roman R. Electrophysiological response in oddball paradigm studied by intracerebral
recordings. PhD thesis in Physiology and Pathological Physiology, Faculty of Medicine,
Masaryk University, Brno, 2004.
Rosburg T, Trautner P, Ludowig E, Schaller C, Kurthen M, Elger CE, Boutros NN.
Hippocampal event-related potentials to tone duration deviance in a passive oddball
paradigm in humans. NeuroImage 2007;37(1):274-281.
Rueda MR, Posner MI, Rothbart MK. Attentional control and self-regulation. In Baumeister
RF, Vohs KD (eds.). Handbook of self-regulation research theory and applications. New
York: Guilford. 2004;pp. 283-300.
Van Veen V, Carter CS. The anterior cingulate as a conflict monitor: fMRI and ERP studies.
Physiol Behav 2002;77:477-482.
Rumelhart DE, McCelland JL, and PDP Research group. Parallel distributed processing:
Exploration in the microstructures of cognition. MIT Press, Cambridge, MA, 1986.
Rutishauser U, Ross IB, Mamelak AN, Schuman EM. Human memory strength is predicted by
theta-frequency phase-locking of single neurons. Nature 2010;464:903-907.
Sahin NT, Pinker S, Cash SS, Schomer D, Halgren E. Sequential processing of lexical,
grammatical, and phonological information within Broca’s area. Science 2009; 326:445-
449.
Salisbury DF, Rutherford B, Shenton ME, McCarley RW. Button-pressing affects P300
amplitude and scalp topography. Clin Neurophysiol 2001;112:1676-1684.
133
Sato W, Kochiyama T, Uono S, Matsuda K, Usui K. Rapid amygdala gamma oscillations in
response to fearful facial expressions. Neuropsychologia 2010;49:612-617.
Sergeant J, Geuze R, Winsum W. Event-Related Desynchronization and P300.
Psychophysiology 1987;24(3):272-277.
Shin J. The interrelationship between movement and cognition: Theta rhythm and the P300
event-related potential. Hippocampus 2011;21(7):744-752.
Sperling MR. Intracranial Electroencephalography. Amsterdam: Elsevier, 1993.
Stejskal L. Mozkové odpovědi na zaměřené události: ERP (Event-related potentials). In:
Stejskal L, editor. Evokované odpovědi a jejich klinické využití. Praha Publishing
1993;355-359.
Stevens AA, Skudlarski P, Gatenbay Jc, Gore JC. Event-related fMRI of auditory and visual
oddball tasks. Mag Reson Imaging 2000;18:495-502.
Strange BA, Dolan RJ. Adaptive anterior hippocampal responses to oddball stimuli.
Hippocampus 2001;11(6):690-698.
Sochůrková D, Brázdil M, Jurák P, Rektor I. P3 and ERD/ERS in a visual oddball paradigm. A
depth EEG study from the mesial temporal structures. J Psychophysiol 2006;20(1):32-39.
Sutton S, Barren M, Zubin J, John ER. Evoked potentials correlates of stimulus uncertainty.
Science 1965;150:1187-1188.
Talairach J, De Ajuriaguerra J, and David M. Études stéréotaxiques et structures encéphaliques
profondes chez l´homme. Presse Méd 1952;28:605-609.
Talairach J, David M, and Tournoux P. L´Exploration Chirurgicale Stéréototaxique du Lobe
Temporal dans L´Epilepsie Temporale Réparage Anatomique Stéréototaxique et
Technique Chirurgicale. Paris: Masson, 1958.
Talairach J, Szikla G, Tourounx P. Atlas d´Anatomie Stéréotaxique du Télencéphale. Paris:
Masson, 1967.
Talairach J, Tournoux P. Co – Planar Stereotaxic Atlas of the Human Brain. Stuttgart, New
York: Georg Thieme, 1988.
Tankus A, Yeshurun Y, Flash T, Fried I. Encoding of speed and direction of movement in the
human supplementary motor area. J Neurosurg 2009;110:1304-1316.
Tucker DM, Roth DL, Bair TB. Functional connections among cortical regions: topography of
EEG coherence. Electroenceph Clin Neurophysiol 1986;63:242-250.
Ulbert I, Halgren E, Heit G, Karmos G. Multiple microelectrode-recording system for human
intracortical applications. J Neurosci Methods 2001a;106:69-79.
Ulbert I, Karmos G, Heit G, Halgren E. Early discrimination of coherent versus incoherent
motion by multiunit and synaptic activity in human putative MT+, Hum Brain Mapp
2001b;13:226-238.
Ullsperger M, von Cramon DY. Subprocesses of performance monitoring: A dissociation of
error processing and response competition revealed by event-related fMRI and ERPs.
NeuroImage 2001;14:1387-1401.
Vanderwolf CH. Hippocampal electrical activity and voluntary movement in the rat.
Electronceph Clin Neurophysiol 1969;26:407-418.
134
Varela FJ, Lachaux JP, Rodriguez E, Martinerie J. The brainweb: Phase synchronization and
large-scale integration. Nat Rev Neurosci 2001;2:229-239.
Verleger R, Neukäter W, Kömpf D, Vieregge P. On the reasons for the delay of P3 latency in
healthy elderly subjects. Electroenceph Clin Neurophysiol 1991;79:488-502.
Vidal F, Hasbroucq T, Bonnet M. Is the “error negativity” specific to errors? Biol Psychol
2000;51:109-128.
Wang HE, Benar CG, Quilichini PP, Friston KJ, Jirsa VK, Bernard C. A systematic framework
for functional connectivity measures. Frontiers in Neuroscience, 2014, vol. 1, no. 22,
doi:10.3389/fnins.2014.00405.
Watrous AJ, Fried I, Ekstrom AD. Behavioral correlates of human hippocampal delta and theta
oscillations during navigation. J Neurophysiol 2011;105(4):1747-1755.
Wessel JR, Danielmeier C, Morton JB, Ullsperger M. Surprise and error: common neuronal
architecture for the processing of errors and novelty. J Neurosci 2012;32(22):7528-7537.
Woods DL, Knight RT, Scabini D. Anatomical substrates of auditory selective attention:
behavioral and electrophysiological effects of temporal and parietal lesions. Cogn Brain
Res 1993;1:227-240.
Wyble BP, Hyman JM, Rossi CA, Hasselmo ME. Analysis of theta power in hippocampal EEG
during bar pressing and running behavior in rats during distinct behavioral contexts.
Hippocampus 2004;14(5):662-674.
Yeung N, Cohen JD, Botvinick MM. The neural basis of error detection: Conflict monitoring
and the error-related negativity. Psychol Rev 2004;111:931- 959.
135
V.I. Seznam recenzovaných prací autora
Recenzované práce v časopisech s impakt faktorem (WOS):
1) Brazdil M, Roman R, Falkenstein M, Daniel P, Jurak P, Rektor I. Error processing evidence
from intracerebral ERP recordings. Experimental Brain Research.
2002;146(4):460-466. IF (200) = 2,300
2) Brazdil M, Roman R, Daniel P, Rektor I. Intracerebral somatosensory event-related
potentials: effect of response type (button pressing versus mental counting) on P3-like
potentials within the human brain. Clinical Neurophysiology. 2003;114(8):1489-1496.
IF (2003) = 2,485
3) Kukleta M, Brazdil A, Roman R, Jurak P. Identical event-related potentials to target and
frequent stimuli of visual oddball task recorded by intracerebral electrodes. Clinical
Neurophysiology. 2003;114(7):1292-1297. IF (2003) = 2,485
4) Brazdil M, Roman R, Daniel P, Rektor I. Intracerebral error-related negativity in a simple
Go/NoGo task. Journal of Psychophysiology. 2005;19(4):244-255. IF (2005) = 0,968
5) Roman R, Brazdil M, Jurak P, Rektor I, Kukleta M. Intracerebral P3-like waveforms and
the length of the stimulus-response interval in a visual oddball paradigm. Clinical
Neurophysiology. 2005;116(1):160-171. IF (2005) = 2,640
6) Brazdil M, Babiloni C, Roman R, et al. Directional Functional Coupling of Cerebral
Rhythms Between Anterior Cingulate and Dorsolateral Prefrontal Areas During Rare
Stimuli: A Directed Transfer Function Analysis of Human Depth EEG Signal. Human
Brain Mapping. 2009;30(1):138-146. IF (2009) = 6,256
7) Brazdil M, Roman R, Urbanek T, et al. Neural correlates of affective picture processing A
depth ERP study. Neuroimage. 2009;47(1):376-383. IF (2009) = 6,132
8) Kukleta M, Bob P, Brazdil M, Roman R, Rektor I. Beta 2-Band Synchronization during a
Visual Oddball Task. Physiological Research. 2009;58(5):725-732. IF (2009) = 1,430
9) Kukleta M, Brazdil M, Roman R, Bob P, Rektor I. Cognitive Network Interactions and
Beta 2 Coherence in Processing Non-Target Stimuli in Visual Oddball Task. Physiological
Research. 2009;58(1):139-148. IF (2009) = 1,430
10) Kukleta M, Bob P, Brazdil M, Roman R, Rektor I. The level of frontal-temporal beta-2
band EEG synchronization distinguishes anterior cingulate cortex from other frontal
regions. Consciousness and Cognition. 2010;19(4):879-886. IF (2010) = 2,027
11) Brazdil M, Janecek J, Klimes P, Marecek R, Roman R, et al. On the Time Course of
Synchronization Patterns of Neuronal Discharges in the Human Brain during Cognitive
Tasks. Plos One. 2013;8(5). IF (2013) = 3,534
12) Roman R, Brazdil M, Chladek J, et al. Hippocampal Negative Event-Related Potential
Recorded in Humans During a Simple Sensorimotor Task Occurs Independently of Motor
Execution. Hippocampus. 2013;23(12):1337-1344. IF (2013) = 4,302
13) Svetlak M, Bob P, Roman R, et al. Stress-Induced Alterations of Left-Right Electrodermal
Activity Coupling Indexed by Pointwise Transinformation. Physiological Research.
2013;62(6):711-719. IF (2013) = 1,487
136
14) Bob P, Roman R, Svetlak M, Kukleta M, Chladek J, Brazdil M. Preictal Dynamics of EEG
Complexity in Intracranially Recorded Epileptic Seizure A Case Report. Medicine.
2014;93(23). IF (2014) = 5,723
15) Brazdil M, Cimbalnik J, Roman R, et al. Impact of cognitive stimulation on ripples within
human epileptic and non-epileptic hippocampus. Bmc Neuroscience. 2015;16.
IF (2015) = 2,304
16) Czekoova K, Shaw DJ, Urbanek T, Chladek J, Lamos M, Roman R, Brazdil M. What's
the meaning of this? A behavioral and neurophysiological investigation into the principles
behind the classification of visual emotional stimuli. Psychophysiology. 2016;53(8):1203-
1216. IF (2016) = 2,668
17) Damborska A, Roman R, Brazdil M, Rektor I, Kukleta M. Post-movement processing in
visual oddball task - Evidence from intracerebral recording. Clinical Neurophysiology.
2016;127(2):1297-1306. IF (2016) = 3,866
18) Kukleta M, Damborska A, Roman R, Rektor I, Brazdil M. The primary motor cortex is
involved in the control of a non-motor cognitive action. Clinical Neurophysiology.
2016;127(2):1547-1550. IF (2016) = 3,866
19) Brazdil M, Pail M, Halamek J, Plesinger F, Cimbalnik J, Roman R, et al. Very HighFrequency
Oscillations: Novel Biomarkers of the Epileptogenic Zone. Annals of
Neurology. 2017;82(2):299-310. IF (2017) = 9,890
Recenzované práce v časopisech bez impakt faktoru (Scopus):
1) Roman R, Brázdil M, Jurák P, Dufek M, Kukleta M. EEG and motor responses to target
stimuli in an oddball paradigm: SEEG recordings in epileptic patients. Homeostasis in
Health and Disease 2000;40(3-4):96-98.
2) Roman R, Brázdil M, Jurák P, Kukleta M, Rektor I. Is the potential evoked by visual
stimulus dependent on verbal instruction, which determined the stimulus significance?
Homeostasis in Health and Disease 2001;41(5):224-226.
3) Kukleta M, Brázdil M, Roman R, Jurák P. Electrophysiological responses to target and
frequent stimuli recorded by intra-cerebral electrodes during visual odd-ball paradigm in
human. Homeostasis in Health and Disease 2001;41(1-2):58-60.
4) Damborská A, Brázdil M, Jurák P, Roman R, Kukleta M. Steep U-shaped EEG potentials
preceding the movement in oddball paradigm: Their role in movement triggering.
Homeostasis in Health and Disease 2001;41(1-2):60-63.
5) Kukleta M, Brázdil M, Roman R, Rektor I. Potentials evoked by non-target stimuli in
visual oddball setup (intra-cerebral study in humans). Homeostasis in Health and Disease
2001;41(5):220-223.
6) Roman R, Brázdil M, Jurák P, Damborská A, Kukleta M. P3 waveform in oddball
paradigm can be time locked to target stimulus or the motor response. Homeostasis in
Health and Disease 2001;41(3-4):115-117.
7) Roman R, Chládek J, Brázdil M, Jurák P, Rektor I, Kukleta M. Evoked EEG responses in
a frequency range of 5.5-15 hz in visual oddball paradigm (SEEG study). Homeostasis in
Health and Disease 2003;42(5):210-212.
137
8) Kukleta M, Brázdil M, Roman R, Rektor I. Temporal characteristics of cortical activation
induced by non-target stimulus in visual oddball setup (intra-cerebral study in humans).
Homeostasis in Health and Disease 2003;42(5):208-209.
9) Kukleta M, Brázdil M, Roman R, Bob P, Rektor I. Spatiotemporal characteristics of EEG
responses to non-target stimuli in oddball task: An exploration in epileptic patients.
Homeostasis in Health and Disease 2005;43(4):212-215.
10) Roman R, Brázdil M, Jurák P, Rektor I, Kukleta M. Time course of EEG responses in
frequency range of 5.5-15 Hz in a visual oddball paradigm (SEEG study). Homeostasis in
Health and Disease 2006;44(1-2):87-89.
11) Kukleta M, Brázdil M, Bob P, Roman R, Rektor I. Synchronization of ECoG activity
during response to non-target stimuli in oddball task: An exploration in epileptic patients.
Homeostasis in Health and Disease 2006;44(4):159-161.
12) Roman R, Chládek J, Brázdil M, Jurák P, Rektor I, Kukleta M. Changes of oscillatory
activity in a visual oddball task (sEEG study). Homeostasis in Health and Disease
2006;44(4):169-171.
13) Damborská A, Brázdil M, Rektor I, Roman R, Kukleta M. Correlation between stimulusresponse
intervals and peak amplitude latencies of visual P3 waves. Homeostasis in Health
and Disease 2006;44(4):165-168.
14) Světlák M, Hodoval R, Damborská A, Pilát M, Roman R, Černík M, Obereignerů R, Bob
P. The emotional impact of the text of cigarette package health warning on older school
age and adolescent children. Cesko-Slovenska Pediatrie 2013;68(2):78-91.
15) Roman R, Světlák M, Damborská A, Kukleta M. Neurophysiology of defence behaviour.
Ceska a Slovenska Psychiatrie 2014;110(2):96-104.