Automatic detection of skin melanoma from images using natural computing approaches

Abstract

Medical imaging is an area of great interest in terms of accuracy, speed and capacity of integration. In order to improve results and ease the physicians' task, some feature enhancement and image processing should be done automatically in order to lead to some features that allow an automatic classification of the images. This paper presents an original approach to construct an automatic melanoma detection system, based on employing natural computing methods for image preprocessing, feature extraction and classification. Among these methods we rely on cellular automata, reaction diffusion cellular neural networks, nonlinear time-series analysis.