Analysis on Skin Lesion Classification Systems and Dermoscopic Feature Analysis for Melanoma

Abstract

Cancer in the recent yeas is one of the significantly complex diseases for diagnostics. The early detection of a deadly form a cancer ie., skin cancer can be done by using image based computer aided diagnostic systems. These systems have a significant potential to screen and detect malignant melanoma. This paper discusses efficient methods using various algorithms for classification of skin cancer. This paper also reviews the different processing steps involved, the various features extracted and classifiers used to classify as melanoma or benign. We also review and examine the current practices, problems and suggest appropriate modeling methodologies.