Preoperative Liver Function Assessment in Biliary Tract Cancer Patients Using Routine Clinical DCE-MRI: A Prospective Study

Motivation: Biliary tract cancers (BTC) requiring extensive liver resections lack sufficient methods of functional evaluation. DCE-MRI derived functional parameters are promising imaging biomarkers for regional liver function evaluation. Goal(s): This study addresses the demand for accurate preoperative functional evaluation and risk assessment in BTC patients undergoing liver resection. Approach: We compared the correlations between functional imaging biomarkers and blood serum markers. By combining functional imaging biomarkers with traditional markers, we built a machine learning model for postoperative complication prediction. Results: We illustrated correlations between functional imaging biomarkers and blood serum markers and the prediction model had an AUC of 0.784. Impact: Our work enhances preoperative risk assessment in biliary tract cancer by integrating MRI-derived functional parameters with serum markers, complementing traditional metrics for more accurate regional liver function evaluation and improving prediction of postoperative complications for personalized surgical planning.