Artificial Neural Network Based Diagnostic System For Melanoma Skin Cancer

Nowadays, skin cancer is a life-threatening disease that causes human death. The human body contains melanocytic cells and these cells are found in the skin. Abnormal growth of melanocytic cells causes skin cancer. This disease can be diagnosed by an expert dermatologist as a result of the interpretation of dermoscopy images by ABCD rule. Because the diagnosis is made by physicians, there are some problems such as misdiagnosis due to human errors. Therefore, in order to solve these problems, computer-assisted diagnosis is necessary to diagnose skin cancer to assist the doctor. The aim of this study was to determine melanoma skin cancer using image processing techniques. In the study, different Artificial Neural Network (ANN) models were applied and their classifier performances were obtained as Multilayer Perceptron (MLP) 99.8 %, Patern Recognition Neetwork 98.3 %, Support Vector Machines 96.7 % and K nearest neighborhood (KNN) 95 %.