A multi-center study in Gut evaluated an AI-based classifier, CompCyst-AI, designed to detect pancreatic cancer in cyst fluid samples by integrating genomic, proteomic, and clinical data. It aims to enhance the risk stratification of pancreatic cystic lesions, which present diagnostic challenges. Conducted on 862 patients undergoing cyst resection, the classifier improved diagnostic accuracy, matching pathology-based management in 72 percent of cases, compared to 52 percent with current methods, and effectively identified high-grade carcinoma while reducing unnecessary surgeries for low-risk lesions.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
