Authors: Om Dwivedi, Neelam Singh Parihar
Abstract: Skin cancer remains one of the most prevalent and life-threatening diseases globally, necessitating early and precise diagnosis. This research proposes an optimized deep learning framework using ResNet152 for automated skin lesion classification. The model integrates preprocessing, segmentation, and feature extraction to enhance lesion detection and classification accuracy. Experimental results demonstrate superior performance, achieving 97% accuracy, 98% precision, and 97% recall, outperforming existing ResNet variants. The framework’s robustness and adaptability make it suitable for clinical and remote diagnostic applications, promoting early intervention and reducing diagnostic errors.
