Detection of Skin Diseases using Resilient Neural Network (RNN)

Dermoscopy is an important tool in the early detection of melanoma, increasing the diagnostic accuracy over clinical visual inspection in the hands of experienced physicians. A pigment network whose structure varies in size and shape is called an irregular or a typical pigment network (APN). Melanoma is one of the lethal skin cancers causing death to thousands of people every year. Early detection of melanoma is possible through visual inspection of pigmented lesions over the skin, treated with simple excision of the cancerous cells. A technique based on CNN (convolution neural network) and FCM (Fuzzy C-mean) clustering for efficient precise and automated Melanoma region segmentation within dermoscopic images. In the proposed system segmentation model is designed henceforth the classification is further done by resilient neural network (RNN) to improve the quality of classification. By using the neural network it will display the types of disease and will determine accuracy increased compared to existing system.