VGG-19-Based Deep Learning for Automated Skin Cancer Detection

Based on the VGG-19 convolutional neural network for the timely identification of skin cancer, this research presents a novel methodology, using images derived from the ISIC archive comprising 1900 normal and 1597 cancerous samples. Recognizing the pivotal role that visual diagnosis plays in detecting malignant transformations in skin cells, the study leverages a deep learning model to facilitate automated classification. The methodology involves key processes such as data loading, categorical labelling, normalization, and model construction using Keras with a TensorFlow backend, all of which comprise a comprehensive 14-step approach. The balanced composition of the dataset ensured rigorous evaluation, ultimately yielding an impressive accuracy of 97.5 \%. This study not only underscores the critical need for early skin cancer identification but also demonstrates the practical application value of the VGG-19 framework in clinical diagnostics, thereby contributing to improved patient outcomes. Moreover, by advancing innovative, AI-driven diagnostic solutions, the research aligns with SDG 3 (Good Health and Well-Being) and SDG 9 (Industry, Innovation, and Infrastructure), underscoring its relevance to sustainable development goals in healthcare and technological innovation.