Artificial Intelligence-Augmented Electrocardiogram Detection of Left Ventricular Systolic Dysfunction in the General...

Area of Science:

• Cardiology
• Medical Diagnostics
• Artificial Intelligence in Medicine

Background:

• Left ventricular systolic dysfunction (LVSD) is a significant cardiovascular condition.
• Early detection of preclinical LVSD is crucial for timely intervention and improved patient outcomes.
• Current screening methods for LVSD may have limitations in community settings.

Purpose of the Study:

• To validate an artificial intelligence-augmented electrocardiogram (AI-ECG) algorithm for detecting preclinical LVSD.
• To assess the performance of AI-ECG in a large, community-based cohort.
• To evaluate the potential of AI-ECG as a screening tool for preclinical LVSD.

Main Methods:

• A cohort of 2041 community-dwelling adults aged 45+ underwent echocardiogram and ECG.
• AI-ECG performance was evaluated for LVSD detection, including preclinical stages after excluding clinical heart failure.
• An imputed screening program and 10-year incident LVSD prediction were modeled.

Main Results:

• AI-ECG demonstrated high accuracy (AUC 0.97) for LVSD detection in the total population and a high-risk subgroup.
• An imputed screening program showed efficiency in identifying preclinical LVSD cases.
• A positive AI-ECG predicted a 2.31-fold increased risk of incident LVSD over 10 years.

Conclusions:

• AI-ECG is a validated tool for identifying preclinical LVSD in a community setting.
• The AI-ECG algorithm shows potential as a cost-effective screening tool for preclinical LVSD.
• Further research is warranted to establish AI-ECG as a routine screening method for preclinical LVSD.