An improved procedure for the automatic detection of dermoscopic structures in digital ELM images of skin lesions

Abstract

Early detection of melanoma is a very critical issue in todaypsilas dermatologic practice. Different diagnostic methods have been proposed which define multiple criteria for the evaluation of the malignancy of a lesion. The paper is devoted to the detection of an important dermatologic structure: the atypical pigmented network. A proposal is described for the application of decision-tree classification techniques to the results of specific image processing algorithms for the estimation of chromatic and structural parameters.