Computer-Assisted Diagnosis of Lymph Node Metastases in Colorectal Cancers
Episode description
Screening lymph nodes for metastases in colorectal cancer (CRC) can be a cumbersome task, but it is amenable to artificial intelligence (AI)-assisted diagnostic solutions. Prof. Inti Zlobec and Dr. Amjad Kahn from the Institute of Pathology in Bern, Switzerland discuss their newly proposed deep learning-based tool for the evaluation of CRC lymph node metastases in digitized whole-slide images. Their approach showed excellent performance, with high sensitivity (0.99) and specificity (0.96) in two validation cohorts of CRC cases (3836 slides) when comparing slide-level labels with the ground truth (pathologist reports). The overlays of AI-based prediction within lymph node regions matched 100% when compared with a microscope evaluation by expert pathologists!