Patch Enhancement for Melanoma Detection With Bag of Visual Words

Abstract

Melanoma is a type of skin cancer caused by the ultraviolet radiation of Sun. Melanoma will become severe if it is not detected early, and it may spread to other body organs, most commonly the lungs, brain, liver, and bones. Dermatologists look for tell-tale signs of melanoma on pigmented skin lesions (moles) to detect it or, in some cases, differentiate it from other skin diseases. Unfortunately, imprecise subjective analysis may result in a series of biopsies that are unnecessary. Furthermore, this type of imprecision may allow a melanoma case to spread undetected. This study develops an automatic melanoma detection system to overcome this challenge. The proposed method is based on Bag of Visual Words (BoVW) with a new patch enhancement scheme, which incorporates both traditional and cutting-edge methods. Experimental comparisons between the proposed method and the well-known convolutional neural network models demonstrate the effectiveness of the developed system.