查看文章

Melanoma is a type of skin cancer, which is not that common like basal cell and squamous carcinoma, but it has dangerous implications since it has the tendency to migrate to other parts of body. So, if it is detected at an early stage then we can easily treat; otherwise it becomes fatal. Many computer-aided diagnostic methods using dermoscopy images have been proposed to assist the clinicians and dermatologists. Along with conventional methods which extract the low level handcrafted features, nowadays researchers have focused towards deep learning techniques which extract the deep and more generic features. Since 2012, deep learning has been applied to classification, segmentation, localization and many other fields and made an impact. This paper reviews about the deep learning techniques to detect melanoma cases from the rest skin lesion in clinical and dermoscopy images.