Conference Coverage

AI-Enabled ECG Accurately Detects Elevated Filling Pressure Among Patients With Hypertrophic Cardiomyopathy

Kate Young

Key Clinical Summary

  • AI-ECG model validated in hypertrophic and amyloid cardiomyopathy, showing high sensitivity for detecting diastolic dysfunction.
  • Accuracy ranged from 65% to 75%, with strongest performance in cardiac amyloidosis.
  • Tool offers a low-cost, scalable option for early identification of elevated filling pressure.

Introduction

Researchers from UZ Leuven (Belgium) and the Mayo Clinic (Rochester, Minnesota) presented new findings at the American Heart Association (AHA) Scientific Sessions demonstrating the diagnostic utility of an artificial intelligence–enabled electrocardiogram (AI-ECG) in detecting diastolic dysfunction among patients with hypertrophic cardiomyopathy (HCM) and cardiac amyloidosis.

The study evaluated how effectively the AI-ECG could predict elevated filling pressures compared with echocardiographic standards.

Study Findings

The multicenter study, led by Robin Van Lerberghe, MD, and colleagues, analyzed 1592 ECGs across several cardiomyopathy phenotypes: non-obstructive HCM (n = 125), obstructive HCM (n = 29), apical HCM (n = 33), and cardiac amyloidosis (n = 49). Echocardiography was performed within 14 days of ECG recording to grade diastolic function using established parameters—E/e’ ratio, left atrial volume index, pulmonary venous atrial reversal flow duration, and peak tricuspid regurgitation velocity.

The AI-enabled ECG, developed at Mayo Clinic, categorized diastolic function as normal (grade 0/1) or abnormal (grade 2/3). Model performance varied by condition:

  • Non-obstructive HCM: 65% accuracy
  • Obstructive HCM: 67% accuracy (sensitivity 94%)
  • Apical HCM: 65% accuracy
  • Cardiac amyloidosis: 75% accuracy (sensitivity 97%)

Systolic dysfunction (left ventricular ejection fraction ≤ 40%) was more prevalent in cardiac amyloidosis (29%) compared with all HCM subtypes (5%). Despite only moderate overall accuracy, the high sensitivity highlights the model’s value in detecting at-risk patients.

Clinical Implications

The study underscores the potential clinical utility of AI-ECG screening for detecting elevated filling pressure—a key feature of heart failure with preserved ejection fraction (HFpEF)—in structurally complex cardiomyopathies. Traditional echocardiographic assessment can be challenging due to variable remodeling patterns in HCM and amyloidosis. The AI-ECG provides a rapid, noninvasive, and widely accessible diagnostic adjunct, potentially improving early disease recognition and optimizing resource allocation, especially in settings with limited echocardiography access. Its scalability and low cost position it as a valuable front-line screening tool to guide further imaging or therapeutic decisions.

Expert Commentary

“The AI-enabled ECG model demonstrates moderate accuracy but high sensitivity for detecting diastolic dysfunction and increased filling pressure, especially in obstructive HCM and cardiac amyloidosis,” the authors concluded. “The AI-enabled ECG is a useful, widely scalable, and low-cost tool to identify patients at increased risk for diastolic dysfunction.”

Conclusion

This collaborative study from UZ Leuven and Mayo Clinic highlights the expanding role of AI-enabled diagnostics in cardiology. By providing sensitive, cost-effective screening for diastolic dysfunction, the AI-ECG may enhance early detection and management of patients with hypertrophic cardiomyopathy and cardiac amyloidosis, potentially improving heart failure outcomes.

Reference:
Van Lerberghe R, Vandenberk B, Jacobs J, et al. Artificial intelligence–enabled electrocardiogram for the detection of elevated filling pressure in hypertrophic cardiomyopathy and cardiac amyloidosis. Presented at: American Heart Association Scientific Sessions; 2025; Chicago, IL.