Abstract
<title>Abstract</title> <p>Melanoma is one of the most aggressive forms of skin cancer and remains a significant global public health challenge due to its high mortality rate when diagnosed at advanced stages. Recent advances in artificial intelligence (AI), particularly deep learning (DL) and convolutional neural networks (CNNs), have transformed automated melanoma detection by improving diagnostic accuracy and supporting clinical decision-making. This study presents a comprehensive bibliometric and systematic literature review of research published between 2015 and 2026 on deep learning-based melanoma detection. Using secondary data collected from peer-reviewed publications indexed in major scientific databases, the review analyzes publication trends, influential authors, datasets, deep learning architectures, evaluation metrics, and emerging research directions. The findings indicate substantial growth in the adoption of transfer learning, explainable AI, ensemble models, and hybrid CNN frameworks. Despite remarkable progress, challenges remain regarding dataset imbalance, model interpretability, clinical validation, and cross-dataset generalization. The study provides researchers with an updated overview of current developments, identifies research gaps, and proposes future research directions for developing reliable and clinically applicable melanoma detection systems.</p>