Fig. 1: Data and workflow for analysis of cancer lymph node metastasis with deep learning.
From: Predicting gastric cancer outcome from resected lymph node histopathology images using deep learning
a Summary of each dataset. b Representative images for each of the five tissue classes we labeled in the lymph node area. c H&E pathological slides were first scanned to obtain WSIs. The WSIs were then labeled for training networks. The trained networks were used to analyze the patient’s WSIs and applied to clinical practice.
