Figure 8

The probability maps obtained via the ConvNetUHCMC/CWRU and ConvNetHUP classifiers on whole-slide images of normal breast sections from the UHCMC/CWRU and NC data cohorts are shown in panels (A–C,D–F) respectively.
The probability maps obtained via the ConvNetUHCMC/CWRU and ConvNetHUP classifiers on whole-slide images of normal breast sections from the UHCMC/CWRU and NC data cohorts are shown in panels (A–C,D–F) respectively.