Motion-corrected echo decorrelation imaging for monitoring exvivo cardiac ablation

Radiofrequency ablation for treatment of cardiac arrhythmia would benefit from real-time monitoring of ablation progress. Echo decorrelation imaging has been shown to predict local ablation but is susceptible to artifacts caused by heart motion. Here, a method for motion-corrected echo decorrelation imaging is tested in ex vivo ablation of bovine cardiac muscle with controlled motion. Experiments employed a Stockert 70 generator and a 4-mm Celsius catheter (Biosense Webster) with the ablation duration of 3min at the target ablation temperature of 90 °C, resulting in thermal lesions of diameter about 1 cm. Cyclical motion was induced by a stepping motor system (Velmex) with velocities of 0–10 mm/s in the travel range of 1.5 cm. Pulse-echo ultrasound images were acquired by an Acuson SC2000 scanner with a Z6Ms transesophageal matrix array (Siemens). Throughout ablation at intervals of 5–6 s, sequentially paired bi-plane images were acquired as complex (IQ) beamformed echoes with an interframe time of 9–13 ms. Regions of interest (ROIs), one including the catheter location and one outside the ablation zone, were tracked using detection of the myocardial wall boundary and spatial cross-correlation. Echo decorrelation within the ablation ROI was then mapped and compensated to remove motion-induced decorrelation measured in the reference ROI. Receiver operating characteristic curves characterize accuracy for predictions of local myocardial ablation by motion-compensated echo decorrelation imaging as a function of the induced motion amplitude.