Skip to main content
Fig. 4 | Journal of Translational Medicine

Fig. 4

From: A novel deep learning-based algorithm combining histopathological features with tissue areas to predict colorectal cancer survival from whole-slide images

Fig. 4

Examples of poor (A1–A5), and better survival cases (B1–B5) and corresponding Kaplan‒Meier survival curves (C1–C5) of tissue area features including max_tumor_area, lymphocyte_inside_tumor, lymphocyte_around_tumor, around_inside_ratio, and total_stroma_area. The cutoff point to categorize tissue area features as high-group and low-group were determined by the maximally selected rank statistics method. There is a whole slide image (left) and a visualized map (right) in each case (A1–A5 and B1–B5). The red outer frame indicates a high-group case (A1, A2, B3, B4, and A5), and the blue outer frame represents a low-group case (B1, B2, A3, A4, and B5). LYM lymphocyte, STR stroma, TUM tumor

Back to article page