Predicting stroke and mortality in mitral stenosis with atrial flutter: A machine learning approach

Area of Science:

• Cardiology
• Medical Informatics
• Machine Learning in Medicine

Background:

• Mitral stenosis (MS) with atrial flutter (AFL) presents complex risks for patients.
• Predicting cerebrovascular events and mortality in this population is clinically significant.
• Existing risk stratification methods may benefit from advanced analytical approaches.

Purpose of the Study:

• To develop and validate a gradient boosting machine (GBM) model for predicting cerebrovascular events and all-cause mortality in patients with MS and AFL.
• To identify key clinical, electrocardiographic, and echocardiographic predictors for these adverse outcomes.

Main Methods:

• A machine learning approach using a gradient boosting machine (GBM) model.
• Analysis of chart data and imaging studies from 2184 patients with MS and AFL.
• Statistical analysis to identify significant risk factors and high-risk features.

Main Results:

• GBM identified significant predictors for transient ischemic attack (TIA)/stroke: mitral valve area (MVA), right ventricular systolic pressure, pulmonary artery pressure (PAP), left ventricular ejection fraction (LVEF), New York Heart Association (NYHA) class, and surgery.
• Predictors for all-cause mortality included MVA, PAP, LVEF, creatinine, hemoglobin, and diastolic blood pressure.
• The model demonstrated effective risk prediction by recognizing key variables.

Conclusions:

• The GBM model effectively integrates diverse clinical data for improved risk prediction in MS with AFL.
• The model enhances the identification of critical variables associated with cerebrovascular events and mortality.
• This approach offers a promising tool for risk stratification and clinical decision-making in this patient cohort.