Personalised thrombo-embolic risk prediction after endometrial cancer surgery: an explainable AI approach using SHAP - PubMed
2 hours ago
- #Endometrial Cancer
- #Risk Prediction
- #AI in Healthcare
- Study develops an explainable machine learning framework for predicting postoperative lower extremity deep vein thrombosis (LEDVT) risk in endometrial cancer patients.
- The Support Vector Machine (SVM) model, using four variables (postoperative D-dimer, age, fibrinogen, clinical stage), showed strong performance with AUCs of 0.828 and 0.819 in internal and external validations.
- SHAP (SHapley Additive exPlanations) was integrated to provide transparent, interpretable risk assessments, highlighting non-linear risk thresholds like those for D-dimer levels.
- A web-based decision support interface was created for real-time, personalized risk prediction to aid in postoperative management.