Temporal stability and specificity of high bipolar electrogram entropy regions in sustained atrial fibrillation: implications for mapping

Dharmaprani, Dhani, McGavigan, Andrew D., Chapman, Darius, Kutlieh, Rayed, Thanigaimani, Shivshankar, Dykes, Lukah, Kalman, Jonathan, Sanders, Prashanthan, Pope, Kenneth, Kuklik, Pawel, and Ganesan, Anand N. (2019) Temporal stability and specificity of high bipolar electrogram entropy regions in sustained atrial fibrillation: implications for mapping. Journal of Electrocardiology, 53. pp. 18-27.

[img] PDF (Published Version) - Published Version
Restricted to Repository staff only

View at Publisher Website: https://doi.org/10.1016/j.jelectrocard.2...
3


Abstract

Background: The potential utility of entropy (En) for atrial fibrillation (AF) mapping has been demonstrated in previous studies by multiple groups, where an association between high bipolar electrogram (EGM) entropy and the pivot of rotors has been shown. Though En is potentially attractive new approach to ablation, no studies have examined its temporal stability and specificity, which are critical to the application of entropy to clinical ablation. In the current study, we sought to objectively measure the temporal stability and specificity of bipolar EGM entropy in medium to long term recordings using three studies: i) a human basket catheter AF study, ii) a tachypaced sheep AF study and iii) a computer simulation study. Objective: To characterize the temporal dynamics and specificity of Approximate, Sample and Shannon entropy (ApEn/SampEn/ShEn) in human (H), sheep (S), and computer simulated AF.

Methods: 64-electrode basket bi-atria sustained AF recordings (H:15 min; S:40 min) were separated into 5 s segments. ShEn/ApEn/SampEn were computed, and co-registered with NavX 3D maps. Temporal stability was determined in terms of: (i) global pattern stability of En and (ii) the relative stability the top 10% of En regions. To provide mechanistic insights into underlying mechanisms, stability characteristics were compared to models depicting various propagation patterns. To verify these results, cross-validation was performed across multiple En algorithms, across species, and compared with dominant frequency (DF) temporal characteristics. The specificity of En was also determined by looking at the association of En to rotors and areas of wave cross propagation.

Results: Episodes of AF were analysed (H:26 epochs, 6040 s; S:15 epochs, 14,160 s). The global pattern of En was temporally unstable (CV- H:13.42% ± 4.58%; S:14.13% ± 8.13%; Friedman- H: p > 0.001; S: p > 0.001). However, within this dynamic flux, the top 10% of ApEn/SampEn/ShEn regions were relatively temporally stable (Kappa >0.6) whilst the top 10% of DF regions were unstable (Kappa <0.06). In simulated AF scenarios, the experimental data were optimally reproduced in the context of an AF pattern with stable rotating waves surrounded by wavelet breakup (Kappa: 0.610; p < 0.0001).

Conclusion: En shows global temporal instability, however within this dynamic flux, the top 10% regions exhibited relative temporal stability. This suggests that high En regions may be an appealing ablation target. Despite this, high En was associated with not just the pivot of rotors but also with areas of cross propagation, which suggests the need for future work before clinical application is possible.

Item ID: 62529
Item Type: Article (Research - C1)
ISSN: 1532-8430
Keywords: Atrial fibrillation, Entropy, Information theory, Mapping, Signal processing
Copyright Information: © 2018 Published by Elsevier Inc
Funders: National Health and Medical Research Council of Australia (NHMRC), National Heart Foundation of Australia (NHF)
Projects and Grants: NHMRC 1063754, NHF 101188
Date Deposited: 15 Mar 2020 23:57
FoR Codes: 11 MEDICAL AND HEALTH SCIENCES > 1102 Cardiovascular Medicine and Haematology > 110201 Cardiology (incl Cardiovascular Diseases) @ 100%
SEO Codes: 92 HEALTH > 9201 Clinical Health (Organs, Diseases and Abnormal Conditions) > 920103 Cardiovascular System and Diseases @ 100%
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page