IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Automatic Hybrid CNN-Based Skin Cancer Classification

Automatic Hybrid CNN-Based Skin Cancer Classification
View Sample PDF
Author(s): Singaravelan Shanmugasundaram (PSR Engineering College, India), Arun Shunmugam D. (PSR Engineering College, India), Anjel Jean Vincy K. (Vel Tech Rengarajan Dr. Sagunthala R&D Institute of Science and Technology, India), Vijaya Rani G. (Francis Xavier Engineering College, India), I. Noormohamed (R.M.K. Engineering College, India), Suresh Chinnathampy M. (Vel Tech High Tech Dr. Rangarajan Dr. Sakunthala Engineering College, India), M. Dhivya (Ramco Institute of Technology, India)and G. Mareeswari (Ramco Institute of Technology, India)
Copyright: 2025
Pages: 36
Source title: Signal and Image Processing Techniques for Defense, Security, and Healthcare
Source Author(s)/Editor(s): B. Omkar Lakshmi Jagan (Vignan's Institute of Information Technology, India), Amrit Mukherjee (University of South Bohemia, Czech Republic), Thayyaba Khatoon Mohammed (Malla Reddy University, India)and Vustikayala Sivakumar Reddy (Malla Reddy University, India)
DOI: 10.4018/979-8-3693-3840-7.ch002

Purchase

View Automatic Hybrid CNN-Based Skin Cancer Classification on the publisher's website for pricing and purchasing information.

Abstract

Skin cancer, particularly dermo-cancer, is a critical health concern with rising incidences worldwide. Automated classification of dermo-cancer from skin images plays a pivotal role in early diagnosis and timely intervention. In this work, hybrid architecture that integrates inception and ResNet models to enhance feature extraction and facilitate hierarchical learning for improved dermo-cancer classification is explored. The inception module contributes to capturing multi-scale features, while the ResNet module addresses the challenges of vanishing gradients and aids in building a more robust and deeper neural network. The proposed hybrid architecture is trained on a comprehensive dataset, and experimental results demonstrate superior performance compared to individual models, achieving enhanced accuracy, sensitivity, and specificity. The approach automated dermo-cancer classification but also holds promise for other medical image tasks, showcasing the potential of hybrid architectures in medical image analysis.

Related Content

R. N. Ravikumar, S. Aarthi, Valisher Sapayev, Alijon Esanov. © 2026. 32 pages.
Md Mehedi Hasan Emon, Tahsina Khan. © 2026. 34 pages.
Zerin Tasnim, Md Mahdi Hasan Ahid, Md. Adnan Rahman, Mohammad Mofasserul Islam, Md. Nafis Fuad, Abu Bakar Abdul Hamid. © 2026. 34 pages.
P. S. Venkateswaran, S. Jeyakumar, S. Devi Kamatchi, S. Manimaran. © 2026. 36 pages.
Aliza, Abdullah, Muhammad Usman. © 2026. 32 pages.
Rohit Yadav. © 2026. 22 pages.
Salam Al E'mari, Yousef Sanjalawe, Fuad Fataftah. © 2026. 30 pages.
Body Bottom