Combined prognostic model outperforms individual models for PDAC survival

A combined prognostic model integrating deep learning-derived digital pathology features and clinical indicators demonstrated enhanced predictive accuracy for pancreatic ductal adenocarcinoma (PDAC) survival. In a study involving 142 postoperative cases, the combined model achieved a Harrell concordance index (c-index) of 0.86 in the training set and 0.77 in the testing set, surpassing the individual clinical (0.76, 0.75) and pathological models (0.82, 0.73). This model shows promise for improved prognostic assessments in PDAC.