Purpose: The use of SABR for treatment of ventricular tachycardia is negatively affected by target motion. The effect of motion can be reduced by treating only during the quiescent interval of the cardiac cycle using a technique such as cardiac synchronized VMAT (CSVMAT). In this work, cardiac magnetic resonance (CMR) cine images for an illustrative patient with impaired left ventricle (LV) function were analyzed to track LV contour point motion and determine an optimum treatment window.
Methods: CMR data included breath-hold scans for 2 short-axis slices and 2-, 3-, and 4-chamber view long-axis slices. Image data and myocardial tissue masks for 30 cardiac cycle phases were converted to endocardial and epicardial contours. Coherent point drift, a non-rigid point set registration algorithm, was used to assign contour point correspondences across all cardiac phases. The motion profile of the target was then determined according to the displacement of the associated contour points. From this temporal motion profile, a CSVMAT or gated treatment window can then be designed to minimize target displacement.
Results: Minimum, mean, and maximum displacements for each segment were displayed in a 17-segment bullseye plot for either the short-axis or long-axis direction. Displacement due to cardiac motion was generally greater towards the basal end of the LV compared to the apex. Choosing basal segment 6 as the target, an optimal beam delivery window of 200 ms (for a 1 s cardiac cycle) reduces the maximum target displacement during delivery from 18.4 mm to 4.2 mm. For mid-cavity segment 12, maximum displacement is reduced from 7.0 mm to 1.2 mm.
Conclusion: A method of estimating LV wall displacement due to cardiac motion was developed. This method may be used to create optimal treatment margin guidelines for cardiac radioablation targeting specific areas of the LV.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by a Vancouver Coastal Health Research Institute (VCHRI) Innovation & Translational Research Award. Authors Steven Thomas , Kirpal Kohli, and Justin Poon have filed a patent application for CSVMAT.