Predictive Approach for Melanoma Skin Cancer Detection using CNN

Abstract

Melanoma skin cancer is one of the most fatal skin cancer forms but can be easily cured if detected early enough. In this paper, we have presented a solution for detecting early stage melanoma skin cancer using a convolution neural network. Image processing is applied on raw images before feeding them to the neural network. We have utilized 514 dermoscopic images from the International Skin Imaging Collaboration (ISIC) archive for training and validating our model. The experimental results obtained via this process were promising and would immensely help dermatologists in early stage detection of melanoma.

Keywords: Melanoma, CNN, Skin Cancer, Image Processing