Machine Learning Approaches in Primary Mitral Regurgitation
- An AI/machine learning (ML) model integrating standard, quantitative, and objective echo parameters has the ability to predict a patient population with primary MR that would benefit from mitral valve surgery.
- These data suggest the potential value of a more robust and global integration of echocardiographic data to enhance risk stratification in patients with primary MR and to potentially guide treatment by determining interventional benefit.
- There is a need for further trials with large, diverse populations and systematic feature selection to validate and improve model prediction and clinical utility.
What is the utility of machine learning (ML) to identify pathophysiologically and prognostically informative primary mitral regurgitation (MR) patient subgroups based on standard echocardiographic measurements?
The investigators used unsupervised and supervised ML and explainable artificial intelligence (AI) to integrate 24 echocardiographic parameters in 400 primary MR subjects from France (n = 243; development cohort) and Canada (n = 157; validation cohort) followed up during a median time of 3.2 (interquartile range [IQR], 1.3-5.3) years and 6.8 (IQR, 4.0-8.5) years, respectively. The authors compared the phenogroups’ incremental prognostic value over conventional MR profiles and for the primary endpoint of all-cause mortality incorporating time-to-mitral valve repair/replacement surgery as a covariate for survival analysis (time-dependent exposure). The association of phenogroups with time-to-event (i.e., death) was examined using Cox-proportional hazard regression analysis.
High-severity (HS) phenogroups from the French cohort (HS, n = 117; low-severity [LS], n = 126) and the Canadian cohort (HS, n = 87; LS, n = 70) showed improved event-free survival in surgical HS subjects over nonsurgical subjects (p = 0.047 and p = 0.020, respectively). A similar benefit of surgery was not seen in the LS phenogroup in both cohorts (p = 0.7 and p = 0.5, respectively). Phenogrouping showed incremental prognostic value in conventionally severe or moderate-severe MR subjects (Harrell C-statistic improvement, p = 0.480; and categorical net reclassification improvement, p = 0.002). Explainable AI specified how each echocardiographic parameter contributed to phenogroup distribution.
The authors report that novel data-driven phenogrouping and explainable AI aided in improved integration of echocardiographic data to identify patients with primary MR and improved event-free survival after mitral valve repair/replacement surgery.
This preliminary study with an AI/ML model integrating standard, quantitative, and objective echocardiographic parameters reports the ability to predict a patient population with primary MR that would benefit from mitral valve surgery and incrementally improved the prognostic value over the conventional classification method in subjects with moderate-severe and severe MR. These data suggest the potential value of a more robust and global integration of echocardiographic data to enhance risk stratification in patients with primary MR and to potentially guide treatment by determining interventional benefit. However, there is a need for further trials with large, diverse populations and systematic feature selection to validate and improve model prediction and clinical utility.
Clinical Topics: Cardiac Surgery, Invasive Cardiovascular Angiography and Intervention, Noninvasive Imaging, Valvular Heart Disease, Cardiac Surgery and VHD, Interventions and Imaging, Interventions and Structural Heart Disease, Echocardiography/Ultrasound, Mitral Regurgitation
Keywords: Cardiac Surgical Procedures, Echocardiography, Heart Valve Diseases, Machine Learning, Mitral Valve Insufficiency
< Back to Listings