Skin Cancer Prediction and Diagnosis Using Convolutional Neural Network (CNN) Deep Learning Algorithm

Abstract

Artificial intelligence (AI) has recently surpassed human performance in several areas; and there is great hope that in the medical field, AI can enable a better prevention, detection, diagnosis and treatment of diseases. Cancer today remains the second leading cause of death. Most, if not all, professionals in the medical field, agree that early detection of cancer offers great chances of healing or disease control. Image processing is a used method used for the detection of skin cancer as soon as the affected area appears. This triggered the idea of creating an Artificial Intelligence; based on Convolution Neural Network (CNN) deep learning algorithm; which allows the diagnosis of skin diseases by simply using an image captured by a mobile application. A mobile application and a web ap- plication were created to help access to this diagnosis in the most landlocked places where the presence of dermatologists is almost non-existent. Through the proposed system skin cancer will predict and identify malignant skin lesions. Dataset from the International Skin Imaging Collaboration (ISIC) Dermoscopic Archive; that support several types of skin cancer; is used to train and test the pro- posed model.