A New Approach for Diagnosis of Melanoma from Dermoscopy Images

Abstract

Melanoma is a potentially life-threatening type of skin cancer, and early detection is crucial for successful treatment. In this study, we propose a novel method for diagnosing melanoma from dermoscopy images. Our approach leverages advanced image processing techniques and an ensemble of deep learners has been developed using learners of VGG, CapsNet, and ResNet to examine dermoscopic images and distinguish between melanoma and benign lesions with high accuracy. Our results indicate that this approach surpasses previous methods in terms of accuracy and has the potential to significantly enhance the diagnosis of melanoma.