Incorporating the Knowledge of Dermatologists to Convolutional Neural Networks for the Diagnosis of Skin Lesions

TL;DR

This paper presents a CNN-based system for skin lesion diagnosis that integrates dermatologists' expert knowledge through specialized network designs and novel blocks, enhancing melanoma detection accuracy.

Contribution

The authors introduce a method to incorporate dermatologist expertise into CNNs for skin lesion classification, including lesion segmentation and structural pattern analysis.

Findings

Improved melanoma detection accuracy over baseline CNN models

Effective integration of lesion segmentation and structural analysis

Novel CNN blocks enhance diagnostic performance

Abstract

This report describes our submission to the ISIC 2017 Challenge in Skin Lesion Analysis Towards Melanoma Detection. We have participated in the Part 3: Lesion Classification with a system for automatic diagnosis of nevus, melanoma and seborrheic keratosis. Our approach aims to incorporate the expert knowledge of dermatologists into the well known framework of Convolutional Neural Networks (CNN), which have shown impressive performance in many visual recognition tasks. In particular, we have designed several networks providing lesion area identification, lesion segmentation into structural patterns and final diagnosis of clinical cases. Furthermore, novel blocks for CNNs have been designed to integrate this information with the diagnosis processing pipeline.