Catheter Ablation -> Ventricular Arrhythmias -> Electrocardiography D-PO06 - Poster Session VI (ID 26) Poster

D-PO06-082 - Computer-based Analysis Of The 12-lead Electrocardiogram To Localize Ventricular Arrhythmia Sources (ID 1355)

  G. Ho: Stock or Stock Options - Non-Public (Including Start-up) - Vektor Medical Inc.


Background: Manual determination of the originating sources of ventricular arrhythmias from the 12-lead ECG is limited by subtle QRS morphology changes and the inability to accurately depict the interpretation on a 3D model of the heart.
Objective: We tested the hypothesis that quantitative 3-dimensional analysis of the ECG may accurately predict source locations of ventricular arrhythmias and pacing.
Methods: Under an IRB-approved protocol, patients presenting for clinically-indicated EPS and ablation were enrolled. 12 lead ECG data were analyzed using computer-based quantitative analysis to localize ventricular activation sources during ventricular tachycardia (VT), premature ventricular complexes (PVCs), and pacing. Clinical source locations were determined using electroanatomic mapping (EAM) utilizing activation mapping and/or pacemapping (Fig 1A). Accuracy was determined by consensus of 3 independent reviewers comparing the EAM results with the 3D model analysis output (Fig 1B). Statistical analysis was performed using McNemar’s test.
Results: In 6 patients (age 67±14, EF 47%±18%), 30 spatially distinct activations patterns (24 pacing sites, 2 VTs, and 4 PVCs) were identified with EAM. The computer-based algorithm accurately mapped 28 out of 30 ventricular sources (93%, p < 0.001) onto a 3D heart model. Analysis required 3±2 minutes per ECG.
Conclusion: Computational 12-lead ECG analysis is feasible and can accurately determine biventricular activation source locations. Additional studies are needed to confirm these results in larger populations.