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

Table 3 Performance comparison via fivefold cross-validation with K-means using the C-index value

Method	WSISA	Histopathological features	Histopathological + tissue area features
LASSO-Cox	0.556 ± 0.073	0.679 ± 0.095	0.694 ± 0.095
RIDGE-Cox	0.620 ± 0.054	0.656 ± 0.027	0.704 ± 0.028
EN-Cox	0.612 ± 0.032	0.651 ± 0.050	0.683 ± 0.043
SSVM	0.603 ± 0.075	0.657 ± 0.048	0.685 ± 0.037
RSF	0.504 ± 0.053	0.615 ± 0.074	0.651 ± 0.046
GBRT	0.498 ± 0.064	0.614 ± 0.036	0.621 ± 0.057

The results highlighted in bold black show the best performance with those methods

ISSN: 1479-5876