An Explainable Deep Learning Approach for Oral Cancer Detection

• Published in: Journal of electrical engineering & technology Vol. 19; no. 3; pp. 1837 - 1848
• Main Authors: Babu, P. Ashok, Rai, Anjani Kumar, Ramesh, Janjhyam Venkata Naga, Nithyasri, A., Sangeetha, S., Kshirsagar, Pravin R., Rajendran, A., Rajaram, A., Dilipkumar, S.
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.03.2024; 대한전기학회
• Subjects: Electrical Engineering; Electrical Machines and Networks; Electronics and Microelectronics; Engineering; Instrumentation; Original Article; Power Electronics; 전기공학; Deep learning; Oral cancer; Transfer learning; Artificial intelligence
• ISSN: 1975-0102; 2093-7423
• DOI: 10.1007/s42835-023-01654-1

With a high death rate, oral cancer is a major worldwide health problem, particularly in low- and middle-income nations. Timely detection and diagnosis are crucial for effective prevention and treatment. To address this challenge, there is a growing need for automated detection systems to aid healthcare professionals. Regular dental examinations play a vital role in early detection. Transfer learning, which leverages knowledge from related domains, can enhance performance in target categories. This study presents a unique approach to the early detection and diagnosis of oral cancer that makes use of the exceptional sensory capabilities of the mouth. Deep neural networks, particularly those based on automated systems, are employed to identify intricate patterns associated with the disease. By combining various transfer learning approaches and conducting comparative analyses, an optimal learning rate is achieved. The categorization analysis of the reference results is presented in detail. Our preliminary findings demonstrate that deep learning effectively addresses this challenging problem, with the Inception-V3 algorithm exhibiting superior accuracy compared to other algorithms.