Multi-channel delineation of intracardiac electrograms for arrhythmia substrate analysis using implicitly regularized convolutional neural network with wide receptive field

HEJC, Jakub, Richard REDINA, Jana KOLAROVA and Zdeněk STÁREK. Multi-channel delineation of intracardiac electrograms for arrhythmia substrate analysis using implicitly regularized convolutional neural network with wide receptive field. Biomedical Signal Processing and Control. London: ELSEVIER SCI LTD, 2024, vol. 94, August 2024, p. 1-18. ISSN 1746-8094. Available from: https://dx.doi.org/10.1016/j.bspc.2024.106274.

Basic information

• Original name: Multi-channel delineation of intracardiac electrograms for arrhythmia substrate analysis using implicitly regularized convolutional neural network with wide receptive field
• Authors: HEJC, Jakub (203 Czech Republic), Richard REDINA (203 Czech Republic), Jana KOLAROVA (203 Czech Republic) and Zdeněk STÁREK (203 Czech Republic, guarantor, belonging to the institution).
• Edition: Biomedical Signal Processing and Control, London, ELSEVIER SCI LTD, 2024, 1746-8094.

Other information

• Original language: English
• Type of outcome: Article in a journal
• Field of Study: 30201 Cardiac and Cardiovascular systems
• Country of publisher: United Kingdom of Great Britain and Northern Ireland
• Confidentiality degree: is not subject to a state or trade secret
• WWW: URL
• Impact factor: Impact factor: 5.100 in 2022
• Organization unit: Faculty of Medicine
• Doi: http://dx.doi.org/10.1016/j.bspc.2024.106274
• UT WoS: 001224357600001
• Keywords in English: Cardiac electrophysiology; Local activation time; Arrhythmia mapping; Convolutional neural network; Wide receptive field; Dilated convolution
• Tags: 14110115, rivok
• Tags: International impact, Reviewed

Abstract

Objective: Automated segmentation of intracardiac electrograms and extraction of fundamental cycle length intervals is crucial for reproducible arrhythmia substrate analysis conducted during electrophysiology procedures. The objective of this study is to develop a robust, computationally efficient end-to-end model for the precise electrogram multi-channel delineation using a highly imbalanced dataset. Methods: A temporal deep convolutional neural network (CNN) based on the UNet architecture incorporating convolutional layers of varying dilation rates was implicitly regularized through data augmentations (DAs), a domain specific Tversky loss function, and distinct labelling strategies for segments comprising atrial fibrillation (AF). An exploratory study utilizing Bayesian search was conducted to optimize architectural and loss function hyperparameters. The impact of dilated convolutions, data augmentations, and labelling strategies on the performance and generalization capability was assessed through an ablation study. The performance of different models was evaluated using a cross-validation procedure and two independent test datasets derived from two separate patient cohorts containing 326, 84, and 97 electrograms encompassing sinus rhythms, abnormal complexes during ongoing tachycardias, and stimulation protocols. Results: A UNet model with optimized loss hyperparameters, a dilated receptive field, and atrial fibrillation (AF) annotated as a positive class (D-UNet-L) achieved an average S & oslash;rensen-Dice coefficient (SDC) of 84.9 % on recordings with regular atrial beats across test datasets, surpassing the performance of models without loss optimization (81.5 %), without dilated kernels (81.3 %), and with inversed AF labelling (77.5 %). Notably, the highest average accuracy (Acc) of 95.8 % for AF recordings was obtained by a model trained on negatively assigned AF segments, outperforming D-UNet-L (88.9 %), the model without loss optimization (81.5 %), and the model without dilated kernels (81.3 %). The reference D-UNet-L model exhibited overall root-mean-square errors of 8.3 and 9.0 ms across test datasets. Additionally, 61.5 % and 20.8 % of delineations exhibited absolute errors below 5 ms and 10 ms, respectively. Disabling data augmentation (DAs) resulted in a 2.7 % decrease in validation SDC and a 5.3 % increase in training SDC." Conclusion: Generalization capability across independent datasets was improved by employing exponentially weighted Tversky loss. The model's segmentation performance on longer sequences with atrial fibrillation was improved by incorporating dilated convolution kernels. Noise-aware and morphology data augmentations effectively mitigated overfitting potential in a limited training dataset. Label noise introduced by annotating atrial fibrillation sequences into a positive class strengthened regularization of the model, particularly in its ability to identify regular beats. However, it also negatively impacted performance on F-waves. Significance: The proposed method diminishes the dependence on manual measurements in complex electrophysiology measurements and enhances utilization of the collected electrophysiology data. The findings underscore the substantial impact of intrinsic regularization techniques and establish guidelines to facilitate methodological decision-making processes.