Artificial intelligence in pancreatic cancer detection: from premise to practice
Misha D P Luyer Lancet Oncol. 2025 Nov 20:S1470-2045(25)00609-6. doi: 10.1016/S1470-2045(25)00609-6. Online ahead of print.
Pancreatic ductal adenocarcinoma (PDAC) remains among the deadliest cancers globally, primarily because symptoms typically appear at a late stage, resulting in diagnosis when curative treatment is no longer possible. Furthermore, detection of PDAC and other periampullary cancers on CT scans is challenging and requires specialised radiological expertise. However, even experienced radiologists miss early-stage tumours on CT scans; in a 2023 study, 15 expert radiologists achieved an average sensitivity of just 79�6% when reading contrast-enhanced CTs for PDAC.1