Machine Learning-Based Discrimination of Cardiovascular Outcomes in Patients With Hypertrophic Cardiomyopathy (JACC Asia May 2024)-Article: Machine Learning-Based Discrimination of Cardiovascular Outcomes in Patients With Hypertrophic Cardiomyopathy

Article: Machine Learning-Based Discrimination of Cardiovascular Outcomes in Patients With Hypertrophic Cardiomyopathy

Pdf Summary

The study aimed to develop and validate machine learning (ML) models for predicting major cardiovascular events in patients with hypertrophic cardiomyopathy (HCM). By analyzing data from two large HCM cohorts from independent referral centers, the logistic regression model demonstrated the best discriminant ability for all outcomes, including all-cause death, heart failure admission, and stroke. The models were found to have comparable or better performance compared to previous studies, with an emphasis on LR model's stability in prediction. The SHAP analysis highlighted the importance of features such as left atrial diameter and hypertension in predicting adverse outcomes in HCM patients. The study's generalizability was noted as a limitation, suggesting the need for validation in diverse ethnic populations and incorporating advanced cardiovascular imaging parameters for further model refinement. The findings underscore the potential of ML models to improve risk stratification and management for high-risk subsets of HCM patients, aiding in early recognition and intervention for adverse cardiovascular events. The study contributes valuable insight into personalized care and outcome prediction for HCM patients.

Keywords

machine learning

cardiovascular events

hypertrophic cardiomyopathy

logistic regression model

all-cause death

heart failure admission

stroke

SHAP analysis

left atrial diameter

hypertension