Skip to main content
. 2022 Jul 18;12:848790. doi: 10.3389/fonc.2022.848790

Figure 3.

Figure 3

The ROC curves of our DCNN model in identifying different breast cancer molecular subtypes on the test set. (A) Classifying luminal A from non-luminal A subtypes among patients older than 50 years. (B) Classifying luminal A from non-luminal A subtypes for tumor sizes ≤20 mm. (C) Classifying triple-negative from non-triple-negative subtypes for patients younger than 50 years. ROC, receiver operating characteristic; AUC, area under the ROC curve; DCNN, deep convolutional neural network.