Measuring myocardial motion changes in tagged MR to establish inverse electro-mechanical coupling for XMR-guided RF ablation

Radio-frequency (RF) ablation uses electrode-catheters to destroy abnormally conducting myocardial areas that lead to potentially lethal tachyarrhythmias. The procedure is normally guided with x-rays (2D), leading to errors in location and excessive radiation exposure. One of our goals is to provide preand intra-operative 3D MR guidance in XMR systems (combined X-ray and MRI room) by locating myocardial regions with abnormal electrical conduction patterns. We address the inverse electro-mechanical relation by using motion in order to infer electrical propagation. For this purpose we define a probabilistic measure of the onset of regional myocardial motion derived from motion fields. The 3D motion fields are obtained using non-rigid registration of tagged MR sequences to track the heart. We also compare regional motion between two different image acquisitions, thus assisting in diagnosing arrhythmia, in follow up of treatment, and particularly in determining whether the electro-physiological intervention succeeded. We validate our methods using an electro-mechanical model of the heart, synthetic data from a cardiac motion simulator for tagged MRI, a cardiac MRI atlas, MRI data from 6 healthy volunteers (one of them subjected to stress), and an MRI study on one patient with tachyarrhythmia, before and after RF ablation. Results seem to corroborate that the ablation had the desired effect of regularising cardiac contraction.