Visual abstract

41379 2013 Article BFmodpathol2013110 Fig1 HTML

Abstract

The detection of metastasis in lymph nodes is a critical step in determining the appropriate course of treatment for cancer patients. While advanced-stage metastases are often visibly evident, early-stage detection remains a challenging task. Traditionally, pathologists have relied on microscopic examination of tissue samples, but advancements in whole-slide imaging technology have made digital pathology increasingly the standard. This shift enables the application of machine learning (ML) models to enhance the accuracy, speed, and consistency of metastasis detection.

Although ML models have demonstrated success in identifying certain cancer types, such as prostate cancer, detecting lymph node metastases presents unique challenges due to the variability in tissue morphology and staining patterns. This project focuses on developing an ML-based model for detecting lymph node metastases in hematoxylin and eosin (H&E)-stained digital tissue samples.

However, due to the limited availability of high-quality annotated data, we propose a multi-step approach rather than tackling the problem directly. This presentation will outline the methodology, the current progress in addressing these challenges, and the future directions for this research.

Slides

Lecture recordings

Readings

• https://doi.org/10.1038/s41467-022-30746-1
  
• https://doi.org/10.1038/s41591-021-01343-4
  
• https://www.tandfonline.com/doi/abs/10.2147/PLMI.S59826
  
• https://arxiv.org/abs/2309.07778
  

Catering

blueberries, pines, apples

Other

Merry Christmas!