Login Search
FMDB Transactions on Sustainable Computing Systems Submit Paper
    Crossref logo

    Archiving

    Portico

    Fernando Martins De Bulhão is a member of

    Crossref logo

End-to-End Machine Learning System for Accurate Melanoma Detection and Risk Assessment

Authors:
J. Angelin Jeba, S. Rubin Bose, R. Regin, A. Gladysmerlin, M. Rehena Sulthana

Addresses:
Department of Electronics and Communication Engineering, S. A. Engineering College, Chennai, Tamil Nadu, India. School of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. Department of Science and Humanities, Dhaanish Ahmed College of Engineering, Chennai, Tamil Nadu, India. Department of Information Technology and Engineering, Melbourne Institute of Technology, Melbourne, Victoria, Australia. 

Abstract:

The Growing incidence of melanoma, the most lethal skin cancer, is one of the motivations for the development of quick and accurate diagnostic techniques. Early diagnosis is key to improving patient survival, but traditional diagnosis can be time-consuming and dependent on specialist availability. This article proposes automated binary classification of skin lesions as malignant (melanoma) or benign using a deep learning approach. Researchers construct a high-quality composite dataset by combining images from multiple public sources, including the HAM10000 dataset and the International Skin Imaging Collaboration (ISIC) archives for 2016-2019. Our model is based on a Convolutional Neural Network (CNN) architecture that uses transfer learning with the EfficientNet model for improved feature extraction. The merged dataset is utilised to train the system to acquire discriminative features of melanocytic lesions. Experimental results demonstrate the model's high efficacy, with excellent accuracy, precision, and recall in dermoscopic image classification. The research demonstrates the power and simplicity of deep learning as a tool to assist dermatologists in accurately and early detecting melanoma, ultimately leading to improved patient outcomes.

Keywords: Skin Cancer; Deep Learning; Convolutional Neural Network (CNN); Transfer Learning; Medical Image Analysis; Dermoscopic Images; Melanocytic Lesions.

Received on: 23/01/2025, Revised on: 19/04/2025, Accepted on: 23/06/2025, Published on: 09/12/2025

DOI: 10.69888/FTSCS.2025.000528

FMDB Transactions on Sustainable Computing Systems, 2025 Vol. 3 No. 4, Pages: 264-277

  • Views : 68
  • Downloads : 14
Download PDF