Automated verification of pulmonary vein isolation in radiofrequency‐ and cryoballoon‐guided ablation. Issue 7 (16th June 2017)

• Record Type: Journal Article
• Authors: De Pooter, Jan; Phlips, Thomas; El Haddad, Milad; Van Heuverswyn, Frederic; Timmers, Liesbeth; Tavernier, René; Knecht, Sebastien; Vandekerckhove, Yves; Duytschaever, Mattias
• Is Part Of: Pacing and clinical electrophysiology. Volume 40:Issue 7(2017)
• Page Start: 779
• Page End: 787
• Publication Date: 2017-06-16
• Subjects: ablation -- algorithm -- atrial fibrillation -- bipolar electrograms -- cryoballoon -- pulmonary vein isolation -- radiofrequency; Cardiac pacing -- Periodicals; Electrophysiology -- Periodicals; 617.4120645
• DOI: 10.1111/pace.13121
• Languages: English
• ISSNs: 0147-8389

Abstract

Abstract: Background: Verification of pulmonary vein isolation (PVI) can be challenging due to the coexistence of pulmonary vein potentials and far‐field potentials. This study aimed to prospectively validate a novel algorithm for automated verification of PVI in radiofrequency (RF)‐guided and cryoballoon (CB)‐guided ablation strategies. Methods: A data set of 620 (RF: 516 EGMs and CB: 104 EGMs) bipolar electrograms (EGM), recorded by circular mapping catheter placed at the left atrium‐pulmonary vein (PV) junction, were prospectively analyzed by a two‐step algorithm. The algorithm differentiates isolated from nonisolated EGMs based on typology and specific parameters of the bipolar EGMs. EGMs were recorded at baseline and after proven isolation in RF‐ and CB‐guided procedures. Additionally, in the RF group, EGMs during encircling of the PVs were analyzed. Results: In the RF and CB group, the algorithm correctly identifies EGMs as isolated or nonisolated with respectively 93% and 96% sensitivity and 86% and 90% specificity. In the RF subgroups of (1) baseline and proven isolated EGMs, (2) EGMs during encircling, and (3) EGMs in redo procedures sensitivity was 96%, 88%, and 100%, respectively, with specificity of 81%, 91%, and 100%. Fourteen out of 14 (100%) reconnected PVs were correctly identified as containing PVPs. Eleven out of 12 (92%) failed freeze attempts were correctly identified as being nonisolated. Conclusion: We validated a two‐step algorithm for automated PVI verification, applicable both for RF‐ and CB‐guided PVI. The algorithm automatically differentiates isolated from nonisolated PVs with high accuracy and without the need for pacing maneuvers.