Validation Of An Algorithm For Automatic Arrhythmia Recognition And 3D Mapping In A Porcine Model

Introduction: Atrial tachycardias (AT) can present multiple sites of origin or circuits which complicates mapping, requiring creation of separate activation maps per site. Objective: To evaluate the Intra-Cardiac Pattern Matching (ICPM) software that automatically detects and assigns different arrhythmia sources to separate 3D activation maps in a porcine model. Methods: To simulate different ATs, continuous pacing at same cycle length was performed from 2-3 right atrial (RA) sites (2 screw-in leads and mapping catheter) for 60-90 seconds before alternating. RA was continuously mapped with a 48-electrode high-density mapping catheter (Octaray). The operator manually switched and added points to the respective maps when the AT changed. Conversely, the ICPM algorithm (Carto Mapping system) automatically assigned each beat to its respective map. Pacing electrodes were repositioned to create a second set of maps. Offline analysis (manual and automatic maps) was performed comparing local activation times (LAT) and mesh coloring values of adjacent points (<5 mm apart). Differences <10 msec were considered a match. Results: Twenty-three different pacing sites were analyzed in 6 swine with 1 manual/1 automatic map per site (46 maps); and 40,176 points were compared (manual and automatic). Individual LATs for manual and automatic maps were compared and matched 91.2% of the time (variance of <10 ms). Mesh coloring values matched using the same criteria. Conclusion: The ICPM algorithm accurately identified changing atrial activation sites and assigned points to appropriate maps >90% of the time compared to manual acquisition.