An in-silico comparison of anatomical and substrate based AF ablation using electro and optical flow mapping

Abstract Background The pursuit of improved outcomes for persistent AF may necessitate the utilization of innovative technologies and ablation methods. We hypothesize that combining electrophysiologic phase and optical flow mapping could offer a novel approach to identify ablation sites in persistent AF while minimizing ablation area, thereby enhancing the effectiveness of substrate-based ablation compared to a purely empiric strategy. Objective To compare the efficacy of a novel electro-optic flow-based ablation strategy (EOF) with empiric ablation for persistent AF patients using a virtual cohort of bi-atrial models. Methods An exploratory cohort (n=20) from a CT-derived bi-atrial in silico population was utilized to simulate AF. A pipeline was developed to assess ablation targets based on simulated AF and to automatically connect targets to closest inert boundaries. As part of target assessment, phase singularities (PS) and Horn-Schunck average optical flow maps were computed. An intersection of information assessed through PS and curl of average optical flow was utilized to generate EOF maps. 5 x strategies comprising purely anatomical, PS based and EOF guided ablations were simulated (1. PVI; 2. PVI+mitral/roof lines; 3. PVI+mitral/box line; 4. PS+anatomical; 5. EOF+anatomical). Ablation efficacy was determined by first examining change of rhythm via ablation and second by calculating the amount of tissue area ablated. Results Ablating PVI alone (Method 1) did not result in AF termination in any case. Ablating PVI with mitral or roof lines (Method 2) resulted in moderate outcomes (6 AF term. / 3 conv. to AT) whereas ablating PVI with mitral and box lines (Method 3) resulted in improved outcomes (9 AF term. / 4 conv. to AT). Compared to purely anatomical approaches, ablating PS or EOF targets and combining them with roof, mitral lines (Method 4 and Method 5 respectively) saw marked improvement (16 AF term. / 2 conv. to AT). Mapping based on EOF (Method 5) resulted in smaller ablation targets compared to a PS guided strategy (Method 4) with the same outcome. Conclusion AF termination to ablated tissue area ratio was higher for EOF based ablation targets compared to PS based or empiric ablations. Algorithmic generation of ablation lesion masks based on predicted targets can provide novel insights into ablation outcomes. Combining EOF targets with empiric lesions represents a potential ablation strategy of individualized substrate modification for persistent AF.Study Pipeline