Standardizing Single-Frame Phase Singularity Identification Algorithms and Parameters in Phase Mapping During Human Atrial Fibrillation

Author(s)

Li, Xin ^{[1, 2]} ; Almeida, Tiago P. ^{[1, 2, 3]} ; Dastagir, Nawshin ^[4] ; Guillem, Maria S. ^[5] ; Salinet, Joao ^[6] ; Chu, Gavin S. ^[1] ; Stafford, Peter J. ^[7] ; Schlindwein, Fernando S. ^{[2, 7]} ; Ng, G. Andre ^{[1, 7]}

Affiliation

^[1] Univ Leicester, Dept Cardiovasc Sci, Leicester, Leics - England

^[2] Univ Leicester, Sch Engn, Leicester, Leics - England

^[3] ITA, Aeronaut Inst Technol, Sao Jose Dos Campos - Brazil

^[4] Univ Auckland, Auckland Bioengn Inst, Auckland - New Zealand

^[5] Univ Politecn Valencia, Valencia - Spain

^[6] Fed Univ ABC, Ctr Engn Modelling & Appl Social Sci, Santo Andre, SP - Brazil

^[7] Glenfield Hosp, Leicester Cardiovasc Biomed Res Ctr, Natl Inst Hlth Rese, Leicester, Leics - England

• Document type: Journal article
• Source: FRONTIERS IN PHYSIOLOGY; v. 11, JUL 21 2020.
• Web of Science Citations: 0

Abstract

Purpose Recent investigations failed to reproduce the positive rotor-guided ablation outcomes shown by initial studies for treating persistent atrial fibrillation (persAF). Phase singularity (PS) is an important feature for AF driver detection, but algorithms for automated PS identification differ. We aim to investigate the performance of four different techniques for automated PS detection. Methods 2048-channel virtual electrogram (VEGM) and electrocardiogram signals were collected for 30 s from 10 patients undergoing persAF ablation. QRST-subtraction was performed and VEGMs were processed using sinusoidal wavelet reconstruction. The phase was obtained using Hilbert transform. PSs were detected using four algorithms: (1) 2D image processing based and neighbor-indexing algorithm; (2) 3D neighbor-indexing algorithm; (3) 2D kernel convolutional algorithm estimating topological charge; (4) topological charge estimation on 3D mesh. PS annotations were compared using the structural similarity index (SSIM) and Pearson's correlation coefficient (CORR). Optimized parameters to improve detection accuracy were found for all four algorithms usingF(beta)score and 10-fold cross-validation compared with manual annotation. Local clustering with density-based spatial clustering of applications with noise (DBSCAN) was proposed to improve algorithms 3 and 4. Results The PS density maps created by each algorithm with default parameters were poorly correlated. Phase gradient threshold and search radius (or kernels) were shown to affect PS detections. The processing times for the algorithms were significantly different (p< 0.0001). TheF(beta)scores for algorithms 1, 2, 3, 3 + DBSCAN, 4 and 4 + DBSCAN were 0.547, 0.645, 0.742, 0.828, 0.656, and 0.831. Algorithm 4 + DBSCAN achieved the best classification performance with acceptable processing time (2.0 +/- 0.3 s). Conclusion AF driver identification is dependent on the PS detection algorithms and their parameters, which could explain some of the inconsistencies in rotor-guided ablation outcomes in different studies. For 3D triangulated meshes, algorithm 4 + DBSCAN with optimal parameters was the best solution for real-time, automated PS detection due to accuracy and speed. Similarly, algorithm 3 + DBSCAN with optimal parameters is preferred for uniform 2D meshes. Such algorithms - and parameters - should be preferred in future clinical studies for identifying AF drivers and minimizing methodological heterogeneities. This would facilitate comparisons in rotor-guided ablation outcomes in future works. (AU)

• FAPESP's process: 17/00319-8 - Atrial substrate identification in patients with chronic atrial fibrillation using multivariate statistical models and multiple attributes from atrial electrograms
• Grantee: Tiago Paggi de Almeida
• Support Opportunities: Scholarships in Brazil - Post-Doctoral