An Improved Framework for Classification of Skin Cancer Lesions using Transfer Learning
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Skin cancer has emerged as a significant global health concern, with a growing incidences worldwide. Timely detection of skin cancer is pivotal, as early diagnosis greatly enhances the prospects of successful treatment. While skin cancer is commonly associated with sun-exposed areas, it can also manifest in less typical regions. Alarmingly, over 123,000 melanoma and 30,00,000 non-melanoma cases are reported annually across the globe. Given the rapid increase in skin cancer cases, early identification of both cancerous and non-cancerous lesions is imperative. In response to the limitations faced by dermatologists in providing consistently accurate diagnoses, there is a pressing need for an automated and efficient diagnostic system. This study presents an innovative and automated approach for skin cancer classification, achieving superior evaluation metrics and enhanced accuracy compared to prior studies and expert dermatologists. Our proposed Convolutional Neural Network (CNN) model, augmented by leveraging pre-trained models, attains an accuracy rate of 94%, with the highest among the pre-trained models reaching 97%. This research offers a promising avenue to revolutionize skin cancer diagnosis, facilitating early intervention and improved patient outcomes.
Fields of Study & Subjects
Cutaneous Melanoma Detection and Management
AI in cancer detection
Artificial Intelligence in Healthcare
Publication details
- Journal / source
- CRC Press eBooks
- Pages
- 140–152
- Publication date
- 27 March 2026
- Language
- EN
- Document type
- Book chapter