Deep learning in the early diagnosis of acute aortic dissection.

Acute aortic dissection (AAD) diagnosis remains perilously delayed, with fatal consequences. This review examines the ability of deep learning (DL) to enhance early detection using routine clinical data, guiding the use of advanced imaging triage. The review discusses diagnostic pitfalls, current decision pathways, and emerging DL models, while addressing critical barriers—data scarcity, class imbalance, and explainability—that must be overcome for safe, effective clinical integration of AI into the AAD diagnostic pathway.