AI analysis of pancreatic tissue predicts postoperative complications

A study involving 134 patients from the RECOPANC trial found that AI-assisted analysis of pancreatic tissue composition can reliably predict postoperative pancreatic fistula (POPF) risk. The strongest determinant identified was the relative fibrotic tissue area, which outperformed conventional surgical assessments of pancreatic texture. These findings highlight the potential of digital pathology to improve risk stratification and outcomes in pancreatic surgery, marking a significant advance in understanding postoperative complications associated with pancreatic resections.