Principles of Computer
Tomography (CT)
Bohatá Šárka
Radiology Department of Faculty
Hospital Brno, Czech Republic
Faculty of Medicine, Masaryk
University Brno
 CT images are acquired while the
x-ray tube i rotating 360dg. around
the patient
 The x-ray beam is collimated in
axial orientation and divergent to
encompass the patient`s width in
the other orientation.
 The intensity of attenuated x-rays
emerging from the patient is
measured with an array of minute
detectors
 The detector array simultaneously
rotates around the patient in unison
with the x-ray tube (third generation
of CT).
 multiple angular views of a
10mm thick slice or less are
obtained, and from numerous
attenuation values measured,
a computer image is
reconstructed using Fourier
transform functions
 it takes 2-3 sec. to acquire an
image ­ or ,,a slice" and 10-15
to process the data for
display.
Generations of CT scanners ­ I.
 First generation used
rotation and translation
movement . There was
one X-ray tube and
one detector only.
 Parallel-beam
 It taked several
minutes to acquire an
image (one slice).
Generations of CT scanners ­ II.
 Second generation used
also rotation and
translation of X-ray tube
and detectors ­ but there
were several detectors
(10-50).
 Small fan-beam
 It taked cca 10-20
seconds to acquire an
image (one slice).
Generations of CT scanners ­ III.
 Third generation use wide row of
detectors (300-600) - large fan-
beam.
 The detector array
simultaneously rotates around
the patient in unison with the xray
tube (without translation
movement).
 The most used type at the
moment.
 It takes cca 0,5-4 seconds to
acquire an image (one slice).
Generations of CT scanners ­ IV.
 Detector ring
 Source-rotation
 Large fan-beam
 Misrepresenting geometry
 Complicated balancing of
rotor
 Multiple images may also be
obtained without ihnterrupting
the x-ray beam between
slices
 A continuous rotation of the
tube and simultaneous
movement of the tabletop
through the gantry
characterize the spiral or the
helical CT
 The slice is controlled by
collimating the beam in axial
direction and may be as thin
as 1mm (for high resolution
imaging)
 CT images are displayed
in a two-dimensional
plane on a 512x512
matrix of image elements
or pixels
 A pixel represents a
tissue volume or voxel
that usually measures
1x1x10mm
 The longest dimension of
voxel is oriented parallel
to the long axis of the
patient and represents
slice thickness
 Each voxel, respecitvely pixel in two­dimensional
imaging, has its proper level of absorption
coefficient ­ it is called Hounsfield unit (HU) (sir
Godfrey Newbold Hounsfield constructed the first
CT)
 Specific shade of gray colour is assigned each
number of HU
Conventional CT scanners
 Employ fan of x-ray beams
and a large detector array
 Involves alternating patient
translation and x-ray
exposure
 Each rotation of x-ray tube
generates data from which
a corresponding axial image
is reconstructed
Helical (spiral) CT
 Slip-ring technology (no electrical cables connecting gantry to ground) allows
source detector assembly to rotate continuously. Previously, frequent, abrupt
changes between scans were necessary to permit winding and unwinding of
cables
 Simultaneous patient translation and x-ray scanning generates volume of
data
 X-ray beam traces a helix of raw data from which axial images must be
generated
 Each rotation generates data specific to an angled plane of section
 To create true axial image, data points above and below desired section must
be interpolated to estimate value in axial plane
 Thus, interval between reconstructed images can be chosen retrospectively
Multislice CT
 Several rows of detectors each above another
 MS- detector array segmented in z axis, a mosaic
 ­Allows for simultaneous acquisition of multiple
images in scan plane with ONE rotation.
 Possibility of isotropic geometric resolution in
coronal and sagital plain (excellent quality of
multiplanar reconstructions).
 X-ray attenuation
values are measured
in Hounsfield units
(HU) and are
represented on the
Hounsfield scale
 water is arbitrarily
assigned a value of
0HU, air -1000HU and
dense cortical bone
+1000HU.
 The human eye is capable of distinguishing only 30-60 levels
of gray ­ when presented with an image of 4000 levels of
gray, the eye cannot perceive such small density differences.
 Two controls provide for selection of window width and
window level. These allow high-contrast viewing around the
area of interest.
 The window width selects only a segment of the Hounsfield
scale and displays that ,,window" at the full 256 gray levels. All
values above the window width are displayed as white and all
values below are diasplayed as black.
bone window- a,
brain window -b
Contrast media application
 intravascular ­ intravenous, intraarterial
(iodine contrast media ­ ionic or non-ionic,
mostly hyperosmolar; nefrotrophic)
 peroral (isodense - water, hypodense -air,
hyperdense ­ iodine or barium enemas)
 intrathecal (liquor spaces, isoosmolar nonionic
high quality contrast media)
 intracavital
Contraindications
 Allergic patient
 Severe renal insufficiency
 (risk of contrast-induced nefropathy)
 Acute stroke
 (dissrupted blood brain barrier)
 Thyreotoxicosis
 Paraproteinemia
 (Bence-Jonesovy protein can preccipitate in renal tubules and
cause acute tubular necrosis)
Complications
 Adverse reaction ­ due to chemotoxicity ­
heat feelings
 Allergic reaction
 Trophic changes in paravascular application
Algorithm of CT examination
1. Topogram
2. Scanning parametres
3. Reconstruction parametres
4. Postprocessing, archiving of the data
Topogram
Range of examination
Slice thickness and number
 Two main algorithms
of reconstruction of the
image:
 Soft tissue (higher
contrast resolution) ­
a,b
 HR ­ high resolution
(superior geometric
resolution ­ c,d)
Postprocessing
CTA
2D oblique view of
cervical spine
CT angiography of
carotic arteries ­
MIP (maximum
intensity
projection)
reconstructions
CTA ­ shadow surface display - SSD
Stenosis qiantiffication CT angiography of
the aorta
CTA - A. renalis duplex dx.,
truncus coeliacomesentericus
V. cava superior sinistra.
3D
reconstruction
of multiple
facial bones
fractures
3D ­
manubrium
sterni
fract.       
Left ­ MIP, right ­ SSD
Volume rendering technique (VRT)
Principles of MR imaging
Bohatá Š.
Magnetic resonance
 Hydrogen nuclei (1H) have a magnetic
behavior
 We use a strong magnetic field (0,2-2T)
 radiofrequency pulses - a resonance of
frequencies - transfer of energy - 1H raising
from lower to higher energy states
 signal from protons = from the pacient body induction
of a current in a receiver coil computer
reconstruction - resulting image
Disadvantages of MRI
 Strong magnetic field! (whole patient is
placed in)
 duration of examination - untill 60 min.
 limited examination space
 the price = availability
 limited examination field (brain + Csp., C+Th,
Th+L)
Benefits of MRI
 Noninvasive technique
 unprecedented soft tissue contrast
 arbitrary plane of cross-section
 MR angiography, ERCP, PMG (without a
contrast medium)
 contrast media - Gd (minimal risk of allergic
reaction)
Contraindications of MRI
ABSOLUTE
 presence of magnetic metallic implants
 pacemaker/ICD
RELATIVE
 claustrophobia
 non-cooperative patient
 I. trimestr of pregnancy
MR principles
 Nuclear magnetic resonance (NMR) is a non-invasive means of
obtaining clinical images and of studying tissue metabolism in
vivo. Bloch and Purcell independently discovered NMR in 1946.
Six years later they were awarded the Nobel Prize for their
achievements. Since then, the development of NMR
spectrometers and NMR scanners has led to the opening up of
whole new branches of physics, chemistry, biology and medicine.
 Nuclei with an odd number of protons and neutrons possess a
property called spin. In quantum mechanics spin is represented by
a magnetic spin quantum number. Spin can be visualised as a
rotating motion of the nucleus about its own axis. As atomic nuclei
are charged, the spinning motion causes a magnetic moment in the
direction of the spin axis. This phenomenon is shown in figure. The
strength of the magnetic moment is a property of the type of
nucleus. Hydrogen nuclei (1H), as well as possessing the strongest
magnetic moment, are in high abundance in biological material.
Consequently hydrogen imaging is the most widely used MRI
procedure.
A charged, spinning
nucleus creates a magnetic
moment which acts like a
bar magnet (dipole).
 The first step in creating a magnetic
resonance image is placing the subject in
a strong magnetic field.
 Presence of a strong magnetic field causes the
nuclear spins of certain atoms within the body,
namely those atoms that have a nuclear spin
dipole moment, to orient themselves with
orientations either parallel or antiparallel to the
main magnetic field (Bo).
 A collection of 1H nuclei (spinning protons) in the
absence of an externally applied magnetic field.
The magnetic moments have random
orientations. (b) An external magnetic field B0 is
applied which causes the nuclei to align
themselves in one of two orientations with
respect to B0 (denoted parallel and anti-parallel).
 The spin axes are not exactly aligned with B0,
they precess around B0 with a characteristic
frequency
 (a) In the presence of an externally applied
magnetic field, B0, nuclei are constrained to
adopt one of two orientations with respect to B0.
As the nuclei possess spin, these orientations
are not exactly at 0 and 180 degrees to B0. (b) A
magnetic moment precessing around B0. Its path
describes the surface of a cone.
 The nuclei precess about Bo with a frequency called the
resonance or a Larmor frequency.
 Because the parallel state is the state of lower energy,
slightly more spins reside in the parallel configuration,
creating a net magnetization represented by a vector M.
A collection of spins at any
given instant in an external
magnetic field, B0. A small net
magnetisation, M, is detectable
in the direction of B0.
 Magnetic resonance occurs when a
radiofrequency pulse , applied at the Larmor
frequency, excites the nuclear spins raising them
from their lower to higher energy states.
Classically this can be represented by a rotation
of the net magnetization,
away from its rest of equilibrium state.
After a 90 degrees RF pulse, M lies in the x-y plane
and rotates about the z-axis. The component of M in
the x-y plane decays over time. An alternating
current is induced in the receiver coil.
 Once the magnetization is deflected, the RF field is switched
of and the magnetization once again freely precesses about
the direction of Bo.
 This time dependent precession will induce a current in a
receiver coil.
 The resultant exponentially decaying voltage, referred to as
the free induction decay (FID), constitutes the MR signal.
After a 90 degrees RF
pulse, M lies in the x-y
plane and rotates about
the z-axis. The
component of M in the
x-y plane decays over
time. An alternating
current, shown in
Figure (b), is induced
in the receiver coil.
 During the period of free precession the
magnetization returns to its original equilibrium
state by a process called relaxation which is
characterized by two time constants, T1 and T2.
T1 and T2 depend on certain
physical and chemical
characteristics unique to tissue
type, therefore contributing
substantially to the capability of
MRI to produce detailed images
of the human body with
unprecedented soft tissue contrast.
Reconstruction of MR image
 How do we make the voxels (volume
elements)?
 How do we know, from which part of the body
cames the signal, how we make the final
image?
 Complex and complicated process, we need
a physical and mathematical device
Slice Selection
 The first step is the selection of a slice, which is achieved by
applying a magnetic field gradient along the z-axis (Gz) during a
90o rf pulse of a specific frequency bandwidth (period 1 of Figure
9.7). When the slice select gradient, Gz, is applied along the zaxis,the
resonance frequencies of the protons become linearly
related to position along the z-axis. Individual resonance
frequencies correspond to individual planes of nuclei.
Frequency Encoding
 After slice selection, the next task is to distinguish signal from
different spatial locations within this slice. This is
accomplished in the x-direction by applying a gradient (Gx),
the frequency-encoding gradient, during the acquisition of the
signal. (time period 3 in Figure 9.11).
The sequence of RF power and
gradient ampitudes used to excite one
slice and encode the positions of the
spins within that slice into the signal.
In this "frequency encoding," the
positions of the spins encoded by
applying a magnetic field gradient in
one of the directions in the excited
slice during the acquisition.
A series of steps illustrating the concept of
frequency encoding to distinguish the signal
coming from two point sources of
magnetization, e.g. small vials of water, in an
object.(left) When no gradient is applied, both
sources of magnetization resonate at the same
frequency and the signal is a simple decaying
sinusoid. When this signal is Fourier
transformed, the signal is shown to contain
only one frequency.(right)
When a gradient is applied, one of the sources
of magnetization precesses at a higher
frequency than the other. The resulting signal
is an interference pattern of the two
frequencies and is shown to contain by
Fourier transformation to contain two distinct
frequencies. Notice that the Fourier
transformed signal is the projection of the
amount of magnetization along the axis along
which the gradient was applied.
That is, in this one dimensional case, the
frequency content of the signal is the image.
Phase Encoding
 Suppose a slice through a homogeneous sample has been
selected and excited as described in Slice Selection section,
and then frequency encoded according to the previous section.
After a short time, the phase of the spins at one end of the
gradient leads those at the other end because they are
precessing faster. If the frequency encoding gradient is
switched off, spins precess (once more) at the same angular
velocity but with a retained phase difference. This phenomenon
is known as phase memory.
 A phase encoding gradient is applied orthogonally to the other
two gradients after slice selection and excitation, but before
frequency encoding. The phase encoding gradient does not
change the frequency of the received signal because it is not
on during signal acquisition. It serves as a phase memory,
remembering relative phase throughout the slice.
Spatial locations of the spins are encoded
into the signal by applying three
orthogonal gradients, techniques that are
called slice selection, frequency encoding,
and phase encoding. In period 1, a 90
degree pulse and a slice selection gradient
excite one slice. In period 2, the initial
frequency encoding gradient and the phase
encoding gradient are applied. In period 3,
a 180 degree pulse is applied, along with a
slice selection pulse (such that only the
spins in the same excited slice are
"flipped,") and in period 4 the frequency
encoding gradient is applied and the signal
is acquired. The sequence shown here is
repeated numerous times (128, 256, 512,
etc. depending on the desired resolution)
each time with a different strength of the
phase encoding gradient.
A complete pulse sequence diagram for
the spin echo sequence.
 Although the signal obtained from one acquisition (slice selection,
phase encoding and frequency encoding) contains information
from all voxels in the imaging slice, the information gathered from
one iteration of this sequence is not sufficient to reconstruct an
image. Consequently, the sequence has to be repeated with
different settings of the phase-encoding gradient Gy.
 To construct a 256 x 256 pixels image a pulse sequence is
repeated 256 times with only the phase encoding gradient
changing.
 A Fourier transformation allows phase information to be extracted
so that a pixel (x, y) in the slice can be assigned the intensity of
signal which has the correct phase and frequency corresponding
to the appropriate volume element. The signal intensity is then
converted to a grey scale to form an image.
T1 weighted image(left) and
T2 weighted image (right),
axial plane ­ left brain
hemisphere edema
Precontrast T1 WI
Postcontrast T1
weighted images
­ cerebral tumor
(glioblastoma)
showing ring
enhancement