To construct and validate a machine learning algorithm for estimating lymph node metastasis (LNM) probability in colorectal cancer (CRC) patients.
Approach:
Predictor Identification: Logistic regression analysis identified six independent predictors: extramural vascular invasion (EMVI), T stage, fibrinogen (FIB), systolic blood pressure (SBP), thrombin time (TT), and alpha-fucosidase (AFU).
Model Development: Various machine learning architectures were generated and assessed using receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA).
Key Findings:
The stochastic gradient boosting (gbm) model showed superior discriminatory power with AUC metrics of 0.815 in the training cohort and 0.733 in the external validation set.
Calibration assessments indicated a strong concordance between projected probabilities and actual observed events.
The xgbTree algorithm also demonstrated commendable predictive efficacy.
