Using Library Resources to Detect Skin Cancer with Artificial Intelligence

Winner of the 2025 Hesburgh Library Sophomore, Junior, or Senior Research Award. Self-directed research project: developing an artificial intelligence model to detect skin cancer with high accuracy—despite starting with no prior coding experience. Motivated by both a passion for science and a desire to test the limits of his independent learning, Michael taught himself Python and advanced AI techniques using the extensive resources provided by the Hesburgh Library. These included textbooks, peer-reviewed research articles, and digital databases such as IEEE Xplore and PubMed. He built a deep learning architecture capable of analyzing complex medical images, achieving an Area Under the Curve (AUC) of over 0.94 and near 90% accuracy on real-world datasets, including one from Memorial Sloan Kettering Cancer Center. The project significantly outperformed dermatologists in several key metrics and is being prepared as a manuscript for publication, with Michael as first author. His work not only underscores the transformative potential of AI in medical diagnostics, but also exemplifies the critical role of academic libraries in democratizing access to research and fostering innovation. The Hesburgh Library provided both the intellectual foundation and the focused physical environment necessary to bring this ambitious project to life—transforming it from an idea into a high-impact diagnostic tool with the potential to improve patient outcomes.