Image Analysis in Dermatology: A Comprehensive Survey of Disease Classification Methods

作者：Yashasvi, Kirana Kumar, Vijilius Helena Raj, A. Sangeetha · 发表于：International Conferences on Contemporary Computing and Informatics · 年份：2024 · DOI：10.1109/ic3i61595.2024.10829266 · 研究领域：Computer Science

One of the most prevalent medical conditions that people have for a long time is skin disease. The diagnosis of skin diseases is primarily based on the doctors’ experience and the results of skin biopsies, which is a laborious procedure. The detection and classification of skin diseases using photographs requires an automated computer-based approach in order to handle the lack of human experts and increase diagnosis accuracy. Accurately classifying skin problems from images is a crucial task that mostly depends on the characteristics of the diseases being studied. The visual characteristics of many skin illnesses are quite similar, making it more difficult to extract relevant elements from the image. Correctly identifying these illnesses from the images would speed up the diagnostic process, increase accuracy, and provide patients with more efficient and affordable care. This study provides an overview of various approaches and strategies for classifying skin diseases, including deep learning-based and conventional, manually constructed feature-based approaches.