Current barriers in using AI to diagnose skin cancers
Harald Kittler, MD, Medical University of Vienna, Vienna, Austria, comments on the current challenges in the clinical implementation of artificial intelligence (AI) for the diagnosis of skin cancer. Key obstacles include determining which situations suit best for current models of AI diagnosis, at which point in the workflow would AI diagnosis be beneficial, and monitoring how users interact with the algorithms. AI models need to be equipped to handle untrained situations to solve complex, downstream tasks beyond a pre-trained set of diagnoses and imitate human learning and adaptability. Randomized trials can help establish which environments are suitable for AI diagnosis in future clinical practice. This interview took place at The European Congress on Dermato-Oncology 2024 in Vienna, Austria. These works are owned by Magdalen Medical Publishing (MMP) and are protected by copyright laws and treaties around the world. All rights are reserved.
