Artificial Intelligence-based Health Information Management System and Key Technology Study of Early Screening and Hierarchical Diagnosis and Treatment of Pancreatic Cancer

What this Trial Is About

Pancreatic ductal adenocarcinoma (PDAC) is a serious cancer with low 5-year survival rates, mainly due to late-stage diagnosis. Researchers are studying an AI-powered screening system that uses non-contrast computed tomography (NCCT) scans taken during routine medical visits or health check-ups. This observational, multi-center study aims to improve early detection of PDAC and its precursor lesions, potentially leading to better patient outcomes and wider access, especially in underserved areas. The study will use a deep learning AI model to automatically detect abnormalities in the pancreas from NCCT images. Cases flagged by the AI will be reviewed by two independent radiologists. If the radiologists suspect cancer or precursor lesions, participants will undergo further testing such as blood tests and advanced imaging. Those diagnosed with PDAC or precursor lesions will receive specialized treatment and consultation, while those without a diagnosis will be followed clinically for at least one year. Participants will be involved through routine CT scans without additional imaging initially. The study will measure the AI system's accuracy by tracking detection rates, recall rates for further testing, and positive predictive value over three years. Secondary outcomes include the proportion of early-stage PDAC detected and survival time over five years. The study involves multiple hospitals and technical support, with ongoing monitoring and clinical follow-up for confirmed or negative cases.

AI-powered Early Detection for Pancreatic Cancer Via Non-contrast CT in Opportunistic Screening Cohort