Automatic planning of atrial fibrillation ablation lines using landmark-constrained nonrigid registration

Abstract

Abstract. Catheter ablation is a common treatment option for drug-refractory atrial fibrillation. In many cases, pulmonary vein isolation is the treatment of choice. With current fluoro overlay methods or electroanatomic mapping systems, it is possible to visualize three-dimensional (3-D) anatomy as well as target ablation lines to provide additional context information. Today, however, these lines need to be set manually before the procedure by the physician, which may interrupt the clinical workflow. As a solution, we present an automatic approach for the planning of ablation target lines. Our method works on surface models extracted from 3-D images. To propose suitable ablation lines, a reference model annotated with reference ablation lines is nonrigidly registered to the model segmented from a new patient's 3-D data. After registration, the reference plan is transferred from the reference anatomy to the individual patient anatomy. Due to the high anatomical variations observed in clinical practice, additional landmark constraints are employed in the registration process to increase the robustness of our approach. We evaluated our method on 43 clinical datasets by benchmarking it against professionally planned ablation lines and achieved an average error over all datasets of 2.7 AE 2.0 mm. A qualitative evaluation of the ablation planning lines matched clinical expectations.

An automatic and accurate registration method for electro‐anatomical map and CT surface

“…Radiofrequency ablation (RF), a technique performed on a 3D electro‐anatomical mapping system, is the most effective treatment for atrial fibrillation (AF) . During the operation, a physician inserts a catheter into the atrial chamber, marks hundreds of mapping points on the atrial wall, and constructs an electro‐anatomical map (EAM) with these points.…”

Optimized viewing angles for cardiac electrophysiology ablation procedures

“…For atrial fibrillation, automatic planning methods have been proposed [11,13]. Based on a particular patient anatomy, e.g.…”