Machine learning-based prediction of ischemic cardio-cerebrovascular events after endovascular or microsurgical treatment of unruptured intracranial aneurysms and risk stratification by the early post-treatment triglyceride-glucose index - Report - MDSpire
Advertisement
Machine learning-based prediction of ischemic cardio-cerebrovascular events after endovascular or microsurgical treatment of unruptured intracranial aneurysms and risk stratification by the early post-treatment triglyceride-glucose index
Clinical Report: Predicting Ischemic Cardio-Cerebrovascular Events Following Treatment of Unruptured Intracranial Aneurysms Using Machine Learning and Early Post-Treatment Triglyceride-Glucose Index for Risk Assessment
Overview
This study establishes a machine learning model to predict ischemic cardio-cerebrovascular events (ICCEs) within 6 months post-treatment for unruptured intracranial aneurysms (UIAs). The early post-treatment triglyceride-glucose (TyG) index is identified as a significant predictor of ICCEs.
Background
Ischemic cardio-cerebrovascular events (ICCEs) are serious complications following treatment for unruptured intracranial aneurysms (UIAs), with an incidence of 5-15%. Effective risk stratification tools are necessary to identify patients at elevated risk for these events post-treatment. Current guidelines primarily focus on rupture risk, leaving a gap in the assessment of ischemic risk.
Data Highlights
Parameter
Value
Patients with ICCEs
240 (12.28%)
Endovascular therapy patients
1,343
Microsurgical intervention patients
611
CatBoost model accuracy
0.875
AUROC
0.945 (95% CI, 0.927–0.963)
HR for TyG index increase
2.61 (95% CI: 2.29–2.96, p < 0.001)
Key Findings
240 out of 1,954 patients (12.28%) experienced ICCEs within 6 months post-treatment.
The CatBoost model achieved an accuracy of 0.875 and an AUROC of 0.945.
Every 1-unit increase in the TyG index correlated with a 2.61-fold increased risk of ICCEs.
The TyG index exhibited a nonlinear relationship with ICCEs, with a threshold effect at approximately TyG = 7.
Machine learning models, particularly CatBoost, outperformed traditional regression models in predicting ICCEs.
Clinical Implications
The early post-treatment TyG index can serve as a practical metabolic marker for assessing ischemic risk in patients with UIAs. Utilizing machine learning models may enhance the predictive accuracy for identifying patients at risk for ICCEs, allowing for more tailored perioperative management.
Conclusion
The study highlights the potential of machine learning and the TyG index in improving risk assessment for ischemic events following UIA treatment, emphasizing the need for advanced predictive tools in clinical practice.
