A review of the quantification and classification of pigmented skin lesions: from dedicated to hand-held devices

Abstract

In recent years, the incidence of skin cancer caseshas risen, worldwide, mainly due to the prolonged exposure toharmful ultraviolet radiation. Concurrently, the computerassistedmedical diagnosis of skin cancer has undergone majoradvances, through an improvement in the instrument and detectiontechnology, and the development of algorithms to processthe information. Moreover, because there has been anincreased need to store medical data, for monitoring, comparativeand assisted-learning purposes, algorithms for data processingand storage have also become more efficient in handlingthe increase of data. In addition, the potential use ofcommon mobile devices to register high-resolution imagesof skin lesions has also fueled the need to create real-timeprocessing algorithms that may provide a likelihood for thedevelopment of malignancy. This last possibility allows evennon-specialists to monitor and follow-up suspected skin cancercases. In this review, we present the major steps in the preprocessing,processing and post-processing of skin lesion images,with a particular emphasis on the quantification andclassification of pigmented skin lesions. We further reviewand outline the future challenges for the creation of minimum-feature,automated and real-time algorithms for the detectionof skin cancer from images acquired via common mobiledevices.