Detection and prognostic stratification of left ventricular systolic dysfunction in left bundle branch block using an...

Area of Science:

• Cardiology
• Artificial Intelligence in Medicine
• Medical Diagnostics

Background:

• Left bundle branch block (LBBB) is associated with an increased risk of left ventricular systolic dysfunction (LVSD) due to cardiac dyssynchrony.
• Artificial intelligence-enabled electrocardiography (AI-ECG) shows potential for LVSD detection, but its efficacy in LBBB patients requires further investigation.
• This study investigated the performance of a validated AI-ECG model in identifying LVSD and predicting outcomes in LBBB patients.

Purpose of the Study:

• To evaluate the accuracy of a previously validated AI-ECG model in detecting LVSD in patients with LBBB.
• To assess the model's ability to predict long-term cardiovascular outcomes in this patient population.
• To determine the clinical utility of AI-ECG for managing LBBB patients with potential LVSD.

Main Methods:

• A retrospective analysis of 5,689 LBBB electrocardiograms (ECGs) from 2,813 patients was conducted using a validated AI-ECG model.
• Left ventricular systolic dysfunction (LVSD) was defined as an ejection fraction ≤40%.
• Model performance was evaluated using AUROC, AUPRC, sensitivity, and specificity. Patients were risk-stratified, and clinical outcomes were compared using Kaplan-Meier analysis.

Main Results:

• The AI-ECG model demonstrated strong performance in detecting LVSD in LBBB patients (AUROC: 0.930, AUPRC: 0.913).
• High-risk patients identified by the model had significantly higher rates of all-cause mortality, device implantation, and cardiovascular hospitalization.
• The model achieved high sensitivity (0.979) but moderate specificity (0.473) for LVSD detection.

Conclusions:

• The AI-ECG model (AiTiALVSD) is effective in diagnosing LVSD in LBBB patients.
• The model successfully stratifies long-term cardiovascular risk, indicating its clinical value for early detection and patient management.
• This AI-ECG tool supports improved clinical decision-making for LBBB patients.