Multi-parametric system for risk stratification in mitral regurgitation: A multi-task Gaussian prediction approach

Area of Science:

• Cardiology
• Medical Informatics
• Biostatistics

Background:

• Mitral regurgitation (MR) risk stratification can be improved with comprehensive data.
• Integrating comorbidities, ECG P-wave indices, echocardiography, neutrophil-to-lymphocyte ratio (NLR), and prognostic nutritional index (PNI) is hypothesized to enhance risk assessment.

Purpose of the Study:

• To evaluate a multi-parametric approach for improved risk stratification in patients with mitral regurgitation.
• To assess the utility of machine learning models in enhancing risk prediction for MR complications.

Main Methods:

• Retrospective analysis of 706 mitral regurgitation patients (2005-2018).
• Inclusion of medical comorbidities, P-wave indices, echocardiographic data, NLR, and PNI.
• Application of logistic regression, decision trees, and a multi-task Gaussian process learning model.

Main Results:

• Age, hypertension, and P-wave duration predicted new-onset atrial fibrillation (AF).
• Low left ventricular ejection fraction (LVEF) and P-wave terminal force predicted transient ischemic attack/stroke.
• Numerous factors including NLR, PNI, and baseline AF predicted all-cause mortality. Multi-task Gaussian process learning showed superior risk stratification.

Conclusions:

• A multi-parametric approach using diverse clinical data significantly enhances risk stratification in MR.
• Multi-task machine learning models offer superior performance for predicting adverse outcomes in MR patients.