Skip to main content
Fig. 2 | Cancer Imaging

Fig. 2

From: Predicting microvascular invasion in hepatocellular carcinoma: a deep learning model validated across hospitals

Fig. 2

A margin of 0.8 produced the best area under the curve (AUC) in the optimization of cropping margin. (A) An illustration of a cropped image (blue square box) with a margin of 0.8 × the edge length (denoted as d) of the labeled bounding box (red square box). (B) A boxplot showing the performance of ResNet-18 model by using images cropped with marginal values ranging from 0.5 to 1.0. A margin of 0.8 yielded the best mean AUC value from the result of a 5-fold cross validation on the training set. The mean AUC is represented by an ‘x’

Back to article page