Classifying Skin Cancer and Acne using CNN

Abstract

Several hospitals and dermatological clinics have adopted computer-vision-based diagnosis tools to aid in the early identification of skin cancer. The most frequent skin diseases are acne vulgaris and skin cancer. Acne Vulgaris affects 85% of the population in their lives, usually during adolescence. Benign Skin Cancer is the cancer commonly affecting people among all other types in developed and developing countries. To measure the success of medical treatment techniques, an objective evaluation of the lesion is required. Traditionally, dermatologists manually count the number of lesions by visual examination or scanning obtained photographs of the patient’s skin and divide them into several categories. This old procedure is time intensive and necessitates a significant amount of work on the part of the physician. Using computer vision, automated the lesion detection, lesion classification, counting of Acne, counting of benign skin cancer and tracking of Acne Severity, making it simple for patients to analyse and track the results of their acne treatment. The goal of this study is to develop a Convolutional Neural network model to classify the lesions into acne and benign skin cancer. The proposed model is developed and trained on acne and different types of benign skin cancer images and achieved an accuracy of 96.4%.