Model-based estimation of AV-nodal refractory period and conduction delay trends from ECG

Area of Science:

• Biomedical Engineering
• Cardiology
• Computational Biology

Background:

• Atrial fibrillation (AF) is a common arrhythmia requiring heart rate management.
• The atrioventricular (AV) node regulates heart rate in AF but often needs pharmacological intervention.
• Current AF treatment selection is empirical, highlighting a need for personalized approaches.

Purpose of the Study:

• To develop a non-invasive method for estimating AV node properties (refractory period and conduction delay) over 24 hours.
• To assess the diurnal and short-term variability of these AV node properties.
• To investigate the predictive value of AV node variability for drug treatment outcomes in AF patients.

Main Methods:

• A novel methodology using a network model, genetic algorithm, and approximate Bayesian computation was employed.
• Non-invasive data from Holter electrocardiograms of 51 permanent AF patients were analyzed.
• Short-term variability was quantified using the Kolmogorov-Smirnov distance between 10-min segments.

Main Results:

• Short-term variability in the fast pathway's refractory period and conduction delay correlated with heart rate reduction during metoprolol treatment.
• No correlation was found between diurnal variability and treatment outcome.
• Machine learning models could not predict drug outcomes based on the analyzed parameters.

Conclusions:

• The developed methodology allows for non-invasive, 24-hour estimation of AV node properties.
• Short-term variability of AV node function shows potential for guiding personalized AF rate-control drug selection.
• Further research is warranted to validate and implement this approach for clinical decision-making.