Obtaining 3D malgnant melanoma indicators through the analysis of surface normals from photometric stereo

Cutaneous Malignant Melanoma (MM) is a life-threatening skin cancer, the successful treatment of which relies heavily on early diagnosis. Overall the last 30 years, the incidence rates of MM have increased more than any other cancer in the UK. The ABCD heuristic rules have been used conventionally as an aid to consultants to differentiate between benign lesions and MMs. However, due to the complexity of the problem caused by large variety of dermatological diseases and various forms of MM, the existing 2D ABCD features have been found to be inadequate in reliably detecting MMs. Motivated by the observations that disruptions in skin texture patterns are larger for MMs than benign lesions, this work investigates the feasibility of lesion classification using 3D skin surface texture features, in the form of 3D skin surface disruptions. This is considered to have the potential of establishing new and independent MM indicators m addition to the existing ID ones.EThOS - Electronic Theses Online ServiceGBUnited Kingdo