ResFCNET: A Skin Lesion Segmentation Method Based on a Deep Residual Fully Convolutional Neural Network

Abstract

Melanoma, a high-level variant of skin cancer, is very difficult to distinguish from other skin cancer types in patients. The presence of a large variety of sizes of lesions, fuzzy boundar ies, their irregular-shaped nature, and low contrast between skin lesions and surrounding flesh areas make it clinically difficult to detect and treat melanoma. In this paper, we propose Residual Full Convolutional Network (ResFCNET), a skin lesion recognition model that com bines residual learning and a full convolutional network to perform semantic segmentation of skin lesion. Based on secondary-feature extraction and classification, an experiment was done to verify the effectiveness of our model using the ISBI 2016 and ISBI 2017 dataset. Results showed that a residual convolution neural network obtained high-precision classification. This technique is novel and provides a compelling insight for medical image segmentation.