Morteza Esmaeili1, Vegard Antun2, Riyas Vettukattil3, Hassan Banitalebi1, Nina Krogh1, and Jonn Terje Geitung1,3
1Akershus University Hospital, Lørenskog, Norway, 2Department of Mathematics, University of Oslo, Oslo, Norway, 3Faculty of Medicine, Institute of Clinical Medicine, University of Oslo, Oslo, Norway
1Akershus University Hospital, Lørenskog, Norway, 2Department of Mathematics, University of Oslo, Oslo, Norway, 3Faculty of Medicine, Institute of Clinical Medicine, University of Oslo, Oslo, Norway
The
explainable method visualized the high-level features of
convolutional neural networks. The method evaluated the performance of deep
learning algorithms on localizing lesions. The proposed training evaluation may
improve human-machine interactions and assist in the training.
Figure 1. Grad-CAM visualizations on tumor detection for different training
networks. The top row depicts the original MR image
examples from four subjects. The magenta counters indicate the tumor lesion
boundaries. The bottom rows show the Grad-CAM
visualizations for three different training algorithm on the selected axial
slices.
Table 1. Mean classification and localization error (%) on the testing
database for DenseNet, GoogleNet, and MobileNet.
