Fig. 1: Overview of the gastric LN assessment workflow. | Nature Communications

Fig. 1: Overview of the gastric LN assessment workflow.

From: Deep neural network trained on gigapixel images improves lymph node metastasis detection in clinical settings

Fig. 1

a Description of the data sets in this study, including the number (P: positive, N: negative) of studies, slides, and LNs; the distributions of age, sex (F: female, M: male), and the Lauren classification (IT: intestinal type, MT: mixed type, DT: diffuse type), the grading of gastric cancer, and AJCC T and N categories. b Pipeline for training the LN detector and the metastasis identification module. c The inference pipeline leverages the trained models to provide prediagnostic predictions of positive and negative LN counts and highlights suspicious areas. d Schematic of the workflow.

Back to article page