Melanoma Detection and Classification based on Dermoscopic Images using Deep Learning Architectures-A Study

Skin cancer is the abnormal growth of skin cells. Melanoma is the very dangerous form of skin cancer, it spreads to neighboring tissue very rapidly. Thus the early detection of melanoma is required. Here we examine existing approaches to the automatic identification and categorization of Melanoma in dermoscopic pictures, emphasizing the major features and main discrepancies between the methodologies used. The main goal is to highlight the benefits and drawbacks of the various approaches. Unlike other studies that just explain and evaluate different approaches qualitatively, this one includes a quantitative comparison. Using distinct skin lesion databases, the performance of numerous melanoma detection algorithms is compared. The accuracy, specificity, and sensitivity of the results are presented.