Biventricular imaging markers to predict outcomes in non-compaction cardiomyopathy: a machine learning study

Area of Science:

• Cardiology
• Medical Imaging
• Machine Learning

Background:

• Left ventricular non-compaction cardiomyopathy (LVNC) is a genetic heart disease presenting with heart failure, arrhythmias, and embolic events.
• Accurate prediction of clinical outcomes in LVNC patients remains a challenge.

Purpose of the Study:

• To identify imaging predictors of major adverse cardiovascular events (MACEs) in LVNC patients using machine learning.
• To analyze echocardiographic (echo) and cardiac magnetic resonance imaging (CMRI) parameters for long-term outcome prediction.

Main Methods:

• Retrospective analysis of 108 LVNC patients diagnosed by echo and CMRI criteria.
• Supervised machine learning applied to a comprehensive set of echo and CMRI parameters.
• Long-term follow-up data including MACEs (death, transplantation, heart failure hospitalization, etc.) were collected.

Main Results:

• Forty-seven percent of patients experienced at least one MACE during a mean follow-up of 5.8 years.
• A combination of four parameters (LV ejection fraction by CMRI, RV end-systolic volume by CMRI, RV systolic dysfunction by echo, and RV lower diameter by CMRI) achieved 75.5% accuracy in predicting MACEs.
• These predictors were effective even in patients with normal LV function and no late gadolinium enhancement.

Conclusions:

• Biventricular assessment is vital for determining LVNC severity and planning clinical interventions.
• Machine learning offers a promising approach for analyzing complex imaging data to stratify risk and predict prognosis in LVNC.
• Early identification of high-risk patients through imaging analysis can improve clinical management of LVNC.