Assessment of the atrial fibrillation burden in Holter electrocardiogram recordings using artificial intelligence

Elisa Hennings; Michael Coslovsky; Rebecca E Paladini; Stefanie Aeschbacher; Sven Knecht; Vincent Schlageter; Philipp Krisai; Patrick Badertscher; Christian Sticherling; Stefan Osswald; Michael Kühne; Christine S Zuern; Swiss-AF Investigators

doi:10.1016/j.cvdhj.2023.01.003

Assessment of the atrial fibrillation burden in Holter electrocardiogram recordings using artificial intelligence

Cardiovasc Digit Health J. 2023 Jan 27;4(2):41-47. doi: 10.1016/j.cvdhj.2023.01.003. eCollection 2023 Apr.

Authors

Elisa Hennings^{1

2}, Michael Coslovsky³, Rebecca E Paladini^{1

2}, Stefanie Aeschbacher^{1

2}, Sven Knecht^{1

2}, Vincent Schlageter¹, Philipp Krisai^{1

2}, Patrick Badertscher^{1

2}, Christian Sticherling^{1

2}, Stefan Osswald^{1

2}, Michael Kühne^{1

2}, Christine S Zuern^{1

2}; Swiss-AF Investigators

Affiliations

¹ Cardiovascular Research Institute Basel, University Hospital Basel, University of Basel, Basel, Switzerland.
² Cardiology, University Hospital Basel, University of Basel, Basel, Switzerland.
³ Department of Clinical Research, University Hospital Basel, University of Basel, Basel, Switzerland.

Abstract

Background: Emerging evidence indicates that a high atrial fibrillation (AF) burden is associated with adverse outcome. However, AF burden is not routinely measured in clinical practice. An artificial intelligence (AI)-based tool could facilitate the assessment of AF burden.

Objective: We aimed to compare the assessment of AF burden performed manually by physicians with that measured by an AI-based tool.

Methods: We analyzed 7-day Holter electrocardiogram (ECG) recordings of AF patients included in the prospective, multicenter Swiss-AF Burden cohort study. AF burden was defined as percentage of time in AF, and was assessed manually by physicians and by an AI-based tool (Cardiomatics, Cracow, Poland). We evaluated the agreement between both techniques by means of Pearson correlation coefficient, linear regression model, and Bland-Altman plot.

Results: We assessed the AF burden in 100 Holter ECG recordings of 82 patients. We identified 53 Holter ECGs with 0% or 100% AF burden, where we found a 100% correlation. For the remaining 47 Holter ECGs with an AF burden between 0.01% and 81.53%, Pearson correlation coefficient was 0.998. The calibration intercept was -0.001 (95% CI -0.008; 0.006), and the calibration slope was 0.975 (95% CI 0.954; 0.995; multiple R² 0.995, residual standard error 0.017). Bland-Altman analysis resulted in a bias of -0.006 (95% limits of agreement -0.042 to 0.030).

Conclusion: The assessment of AF burden with an AI-based tool provided very similar results compared to manual assessment. An AI-based tool may therefore be an accurate and efficient option for the assessment of AF burden.

Keywords: Artificial intelligence; Atrial fibrillation; Atrial fibrillation burden; Deep learning; Holter ECG; Machine learning.