Machine Learning Models to Predict Laparoscopic Surgery Difficulty in Rectal Cancer

Researchers developed and validated machine learning (ML) models to predict the difficulty of laparoscopic total mesorectal excision (LaTME) in rectal cancer. In a study with 186 patients, four ML models—logistic regression (LR), support vector machine (SVM), random forest (RF), and decision tree (DT)—were developed using preoperative and intraoperative factors. All models showed strong performance in predicting surgical difficulty, with SVM achieving the highest area under the receiver operating characteristic curve (AUC = 0.987). A logistic regression-based nomogram was also created, incorporating key predictive factors like body mass index (BMI), tumor diameter, and comorbid hypertension, providing surgeons with a visual tool to assess surgical difficulty.