Fuzzy Deep Learning in Skin Cancer Diagnosis: A Review

Skin cancer(SC) is one of the fatal and prevalent malignancies that cause many deaths every year. There are various causes for SC like exposure to ultraviolet radiation, or genetic factors. It is important to diagnose and detect SC at an early stage to reduce fatalities. In this direction, Researchers have developed multiple approaches for early diagnosis of SC. The advancement in computational resources and deep learning (DL) has improved medical image analysis and aided in the early diagnosis of several diseases. However, traditional DL models are black boxes that lack explainability and are inefficient in handling imprecision, ambiguity and vagueness in real-life problems. Researchers have proposed fuzzy deep learning (FDL) approaches to overcome the limitations of traditional DL approaches in medical image analysis. This paper reviews FDL approaches for early diagnosis of SC. It presents a brief overview of fuzzy logic and DL techniques discussing their advantage and usefulness. It also presents a general framework for SC diagnosis using FDL. Furthermore, it discusses the limitations and future research perspectives in SC diagnosis using FDL. It is believed that the researchers working in the relevant area will find this paper highly beneficial.