Machine learning predicts prolonged surgery in laparoscopic cholecystectomy

A recent study revealed that machine learning can effectively predict prolonged operative times in fluorescent laparoscopic cholecystectomy, identifying 29% of patients at risk. Key predictive factors included type of cholecystitis, puncture ports, gallbladder adhesion, pre-surgery antibiotics, and gallbladder thickness. Using the light gradient boosting machine (LightGBM) model, researchers achieved an impressive AUC of 0.876, indicating strong model performance and clinical utility for surgeons in assessing operative time risks.