Computer-Assisted Diagnosis of Skin Cancer at Early Stage Using Deep Learning Algorithms

Skin is largest organ of body, and it contains 3 layers: epidermis (epion), dermis (layer of tissue), hypodermis (hypo-below). Cancer is abnormal multiplication of cells. If the organ affected is skin then called skin cancer. The evolution of modern technology like computers, AI made skin disease detection easy, quick, and cheap. These deep learning algorithms help in image classification and extraction of features of the digital image. Our approach being simple: we take digital image of affected area then analyze the image and extract features using RCNN (train then test dataset). It is simple and cheap. It requires only a camera and a computer. Since end of twentieth century, the advancements in area of diagnosis and research of skin disease became rapid. The advancement in technology promoted the early diagnosis of skin disease. We are able to detect seven types of skin cancer and get accuracy of 97.83%. The skin cancer is also classified on the basis of gender, age, location of cancer, and the frequency of each cancer (max occurrence) is also calculated.