Spatial variability of catheter positions affects omnipolar mapping in 2D atrial sheet simulations

The omnipolar mapping technique (OT) has recently emerged as a methodology to overcome the sensitivity of bipolar recordings to catheter orientation. It relies on catheters with electrodes typically arranged in square or triangular geometries. Assessing whether OT can be applied in sequential mapping without the use of specialized geometries is still a matter of investigation. In this study, wetackled this issue by exploring the use of OT with multiple bipolar recordings placed at random positions around an anatomical point of interest, without requiring specialized catheter designs. In this way we modeled spatial variability occurring on a beat-to-beat basis. We simulated a slab of atrial tissue with different conduction velocities by solving the bidomain equations using openCARP. We applied OT to the synthetic data while varying the number M = 3,5,10,20 of available bipoles placed at random within radii r = 0.5,1,2 mm around points of interest selected in the mesh. Angles of wave propagation and omnipolar voltages were compared to ground-truth values obtained by local activation time and bipolar voltage maps. Results showed that angles were well estimated when M increased and r decreased, reaching errors between 3◦ to 30◦. Voltage displayed promising results for r = 1 mm.