A Deep Learning Based Critical Analysis Of Skin Lesion Classification

In everyday life, various exposures at work or at home can lead to skin conditions such as allergies or infections. Skin lesions serve as essential indications, alerting to future issues and requiring immediate care and attention. To search for publications concerning skin lesions and diverse procedures, the most well-known scientific databases, such as web of science, scopus, and IEEE Xplore, were used. To find comparable documents, search engines use keywords such as “skin lesion,” “CNN,” “skin diseases,” and so on. Following a comprehensive screening method, 658 abstracts from databases are collected, and 19 full-text articles are identified for consideration. This paper investigates the profound impact of machine learning in dermatology, with a focus on skin lesion classification. This study provides a critical examination of CNN-based models, emphasizing their importance and difficulties in increasing precision diagnosis and treatment in dermatology. Overall, this study provides helpful insights into various approaches for skin lesion classification.