Using Machine Learning to Predict Outcomes Following Transfemoral Carotid Artery Stenting.

Area of Science:

• Cardiovascular Medicine
• Medical Informatics
• Machine Learning in Healthcare

Background:

• Transfemoral carotid artery stenting (TFCAS) is associated with significant perioperative risks.
• Existing outcome prediction tools for TFCAS are limited in their accuracy.
• There is a need for improved methods to predict post-TFCAS complications.

Purpose of the Study:

• To develop and evaluate machine learning (ML) algorithms for predicting 1-year stroke or death after TFCAS.
• To compare the performance of ML models against traditional logistic regression.

Main Methods:

• Utilized the Vascular Quality Initiative (VQI) database (2005-2024) with 35,214 TFCAS patients.
• Extracted 112 features (preoperative, intraoperative, postoperative).
• Trained six ML models, including extreme gradient boosting (XGBoost), using a 70/30 train/test split and 10-fold cross-validation.

Main Results:

• The best ML model (XGBoost) achieved an Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.94 for preoperative prediction.
• Model performance remained high across stages: AUROC 0.94 (intraoperative), 0.98 (postoperative).
• XGBoost significantly outperformed logistic regression (AUROC 0.65).

Conclusions:

• Machine learning models, particularly XGBoost, demonstrate high accuracy in predicting 1-year stroke or death post-TFCAS.
• ML offers a superior alternative to logistic regression for TFCAS outcome prediction.
• These findings can enhance clinical decision-making and patient risk stratification.