ABSTRACT

Design and implementation of a medical image processing system that will provide decision support to the clinician in the diagnosis of melanoma type skin cancers by performing the analysis of dermoscopic images.

Visual features of pigmented lesions are converted into measurable numerical quantities by employing digital image processing methods and a classification regarding melanoma diagnosis is performed based on these quantitative data.

We achieved numerical results showing asymmetry, border and color features of the pigmentary lesions by using segmentation, image histogram, thresholding, convex hull, color clustering, color quantization and distribution methods.

The system under development speeds up the decision process of the clinician. In addition, it allows the diagnosis to be based on more objective data.