Open AccessDetection of Melanoma Skin Cancer with Deep Neural Networks
Publication year - 2019
Publication title -
medical and clinical research
Language(s) - English
Resource type - Journals
ISSN - 2577-8005
DOI - 10.33140/mcr.04.04.05
Subject(s) - convolutional neural network , skin cancer , artificial intelligence , computer science , melanoma , deep learning , deep neural networks , pattern recognition (psychology) , test set , artificial neural network , contextual image classification , set (abstract data type) , dermatology , cancer , medicine , image (mathematics) , cancer research , programming language
Detection of skin cancer involves several steps of examinations first being visual diagnosis that is followed bydermoscopic analysis, a biopsy, and histopathological examination. The classification of skin lesions in the first stepis critical and challenging as classes vary by minute appearance in skin lesions. Deep convolutional neural networks(CNNs) have great potential in multicategory image-based classification by considering coarse-to-fine image features.This study aims to demonstrate how to classify skin lesions, in particular, melanoma, using CNN trained on data setswith disease labels. We developed and trained our own CNN model using a subset of the images from InternationalSkin Imaging Collaboration (ISIC) Dermoscopic Archive. To test the performance of the proposed model, we useda different subset of images from the same archive as the test set. Our model is trained to classify images into twocategories: malignant melanoma and nevus and is shown to achieve excellent classification results with high testaccuracy (91.16%) and high performance as measured by various metrics. Our study demonstrated the potential ofusing deep neural networks to assist early detection of melanoma and thereby improve the patient survival rate fromthis aggressive skin cancer.