Evaluation of different mapping techniques for the integration of electro-anatomical voltage and imaging data of the left ventricle

Abstract

Integration of electrical and structural information about substrate in the left ventricle is very important to guide ablation therapies in ventricular tachycardia cases. This integration asks for finding a mapping between electro-anatomical voltage mapping and delay-enhancement magnetic resonance images. We present an evaluation of the accuracy of some mapping strategies, including different standard rigid and non-rigid registration techniques. We also developed a new mapping algorithm to be applied once both geometries are roughly aligned to improve the currently used simple closest point projection. The new mapping algorithm is based on establishing a homeomorphism between both surfaces using a common surface parametrization computed by mesh flattening, then preserving all original information in both modalities. We applied the different mapping strategies to clinical and synthetic data. Results demonstrated a substantial reduction of the point-to-surface error when using the non-rigid registration technique and an improved substrate overlap when using the proposed mapping algorithm. © 2013 Springer-Verlag.