Background: Mapping regular rhythms with varying cycle lengths (CLs) can be challenging and time consuming when each rhythm must be independently mapped. SuperMap, a novel multi-position non-contact (MPNC) mapping method, automatically differentiates and groups unique rhythms according to unipolar beat morphology and creates a high- resolution activation map for each group.
Objective: Evaluate SuperMap accuracy in grouping unique rhythms in silico and in vivo.
Methods: Focal activations were simulated from 15 locations on a left atrial (LA) model. Twelve electrode locations were used to simulate unipolar CS EGMs. Simulations were repeated varying the number of CS electrodes (4, 8 and 12) and active focal sites (2, 3 and 4). For each of these 9 combinations, the active focal sites were randomized 25 times to produce varying CS beat morphologies (N=225) to test grouping accuracy. In a preclinical study, a continuous recording with mixed CS distal pacing (CSd), proximal pacing (CSp) and sinus (SR) was acquired. In vivo clinical recordings (N=7) were collected during CSd and CSp pacing in 3 patients.
The grouping accuracy was evaluated by F-score, F = 2TP / (2TP + FP + FN).
Results: The mean grouping accuracy in silico was F=0.977. Preclinical in vivo mean grouping accuracy for SR, CSd, and CSp is F=0.949. Clinical in vivo mean accuracy is F=0.983.
Conclusion: SuperMap demonstrates the ability to differentiate and group unique rhythms with high accuracy based on differences in CS beat morphology.