Deciphering atrial repolarization morphology: A spline interpolation framework for atrial arrhythmia diagnosis

Abstract

Background

The characterization of atrial repolarization (Ta wave) remains largely elusive due to its inherently low amplitude and concealment beneath the dominant QRS complex.

Objective

This study aims to witness Ta wave within QRS complex using spline interpolation framework.

Methodology

10-s ECGs of 50 Sinus Tachycardia (SiT) and 20 Atrial Tachycardia (AT) were recorded using standard 12-lead. Lead-II signals were pre-processed for noise removal and fiducial points detection. Later, three spline models were used to synthesize hidden Ta wave using the datapoints from PR and ST segment. Further, validation analysis was performed to select the optimal spline model with the Ta wave of Atrio-Ventricular Block (AVB) ECG.

Results

It was noted that the clamped cubic & B-spline interpolation model gave the best SSIM score of 0.7 and lowest power spectrum % difference of 1.33 of interpolated Ta wave within QRS complex. Further, Ta wave voltage and temporal features including Ta dispersion, area, peak location, Ta area/duration, duration/amplitude, and Ta2/Ta1 were crafted. Statistically significant P, Ta & P-Ta features were fed to seven Machine Learning (ML) models. The best ML models, were used to design a stacked ensemble architecture with combined P-Ta features to enhance the classification accuracy 99% and F1 score 0.99.

Conclusion

The proposed method demonstrated that along with the existing P wave features, Ta wave features have potential in better classification of atrial arrhythmia, while interpolation model offers ease of implementation and adaptability to diverse clinical applications.