Skin Cancer Detection using Deep Learning

Skin Cancer Detection using Deep Learning

Author Name : Mrs. Vijayalaxmi Joshi, Chirag L, Chiranjeevi S, Harsha A

DOI: https://doi.org/10.56025/IJARESM.2025.130125022

ABSTRACT Skin cancer continues to be one of the most common types of cancer worldwide, and early detection is crucial for effective treatment and improved patient outcomes. Advances in medical imaging and deep learning are providing new ways to improve diagnostic procedures, increasing accuracy and efficiency. This research focuses on the development of deep learning-based skin cancer detection algorithms using convolutional neural networks (CNNs) and pre-trained models such as VGG16 and DenseNet. The system classifies dermatoscopic images as benign or malignant using state-of-the-art image preprocessing, feature extraction, and machine learning to optimize performance. High-resolution dermatoscopic images were enriched and normalized to increase the diversity and quality of the training dataset. CNN architecture is used for feature extraction, while VGG16 and DenseNet architectures are optimized to achieve better classification. When combined, the learning process further improves the prediction by combining the performance of individual models. The high accuracy of the system has demonstrated its potential as a reliable tool for early cancer diagnosis. In addition to being efficient, the system also provides doctors with a user-friendly interface for uploading images and getting instant search results. This paper also discusses the challenges encountered, including data limitations and computational requirements, and strategies to overcome these challenges.