Visualizing Convolutional Neural Networks to Improve Decision Support\n for Skin Lesion Classification

Le résumé fourni par la source

Because of their state-of-the-art performance in computer vision, CNNs are\nbecoming increasingly popular in a variety of fields, including medicine.\nHowever, as neural networks are black box function approximators, it is\ndifficult, if not impossible, for a medical expert to reason about their\noutput. This could potentially result in the expert distrusting the network\nwhen he or she does not agree with its output. In such a case, explaining why\nthe CNN makes a certain decision becomes valuable information. In this paper,\nwe try to open the black box of the CNN by inspecting and visualizing the\nlearned feature maps, in the field of dermatology. We show that, to some\nextent, CNNs focus on features similar to those used by dermatologists to make\na diagnosis. However, more research is required for fully explaining their\noutput.\n

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.