Session SA2.1

Coronary Sinus Lead Tracking for Its 3D Dynamic Position Assessment in Cardiac Resynchronization Therapy

C Corsi*, F Veronesi, M Mosconi, C Tomasi,
S Severi, C Lamberti

Universidad de Bologna
Bologna, Italy

Background: Cardiac resynchronization therapy (CRT) has been successfully applied in patients affected by heart failure, left ventricular systolic dysfunction and inter-ventricular conduction delays. However, a variable number of patients (20%-30%) do not respond favorably to CRT and the reasons are not fully understood. One of the possible explanations is a displacement of the stimulation site. To our knowledge no studies have investigated the coronary sinus (CS) lead dislodgement as a possible cause for non-responders. In this study we (1) developed a procedure aimed to quantify the CS lead position in the 3D domain throughout the cardiac cycle and (2) tested this procedure by comparing the CS lead position at device implantation (t0) and at three-months follow-up (t1) using chest fluoroscopy (AP 30° oblique views).
Methods: We defined a protocol to reproduce X-ray data acquisition at t0 and t1 using a radio opaque tattoo and a grid. For each dynamic acquisition, in both views, the catheter tip was manually selected in one frame and automatically tracked applying region matching techniques. This resulted in the tip coordinates time course throughout several cardiac cycles. We filtered these signals from fluctuations due to respiration and synchronized the data obtained from the two views. One cardiac cycle was manually selected in each view; the dynamic tip coordinates and the 3D trajectory throughout the cardiac cycle were reconstructed using stereo-photogrammetric rules. The procedure accuracy was assessed in ten patients by quantifying the distances between the 3D trajectories of the catheter tip in consecutive cardiac cycles. The resolution of our technique was defined as the mean accuracy plus 2 standard deviations. The amplitude of the lead displacement (d) between t0 and t1 was evaluated and the repeatability of its estimate was tested by repeating the analysis three times for different cardiac cycles in six patients for which the follow up acquisitions were available. Depending on d values, patients were categorized in four classes: 0 = no change; 1 = d = 5mm; 2 = d >5mm; 3 - tip completely dislodged from the CS.
Results: Our tracking procedure was successfully applied to ten patients. The accuracy resulted in 0.8±0.5 mm, giving a resolution of 1.8 mm. The repeatability was 0.4±0.4 mm. The six patients in which we evaluated the catheter tip displacement at follow up belonged to class 1 (4) and class 3 (2). These quantitative results were confirmed by the qualitative evaluation of the fluoroscopy videos by the cardiologist.
Conclusions: This new methodology allows quantitative evaluation of CS lead position in 3D space. Measurements were accurate and reproducible. Future correlation, on a larger population, between displacement and electrical parameters could provide both additional explanation for non–responders patients and the basis for new implantation indications.

(Abstract Control Number: 83)