Click here to

Session: Image Registration [Return to Session]

Define Target Volume for Cardiac Radioablation Surgery Using Integration of Endoventricle Electroanatomic Mapping and Coherent Point Drift

T Chen*, H Wang, Y Yuan, C Barbhaiya, D Barbee, New York University, New York, NY


MO-H345-IePD-F1-4 (Monday, 7/11/2022) 3:45 PM - 4:15 PM [Eastern Time (GMT-4)]

Exhibit Hall | Forum 1

Purpose: Cardiac Radiation Ablation (CRA) is an emerging therapy for reducing recurrent Ventricular Tachycardia (VT). Electroanatomic Maps (EAM) delineating important details of the VT substrate are created during catheter ablation, but its integration into standard radiation treatment planning software (TPS) for target definition was limited. Coherent point drift (CPD) provided a robust solution to register two sets of 3D points and was used to define target volumes in the TPS using EAM information.

Methods: The EAM voltage map file was imported into Matlab using in-house script as a set of 3D coordinates of sampling points on the myocardium and EAM voltage at each point. Endocardial surface was also delineated in TPS based on a gated cardiac MR image set acquired using SPGR sequence and contrast agent of 10ml Gadavist, and exported as 3D surface points. Two sets of points were rigidly registered in Matlab using CPD to establish point-to-point correspondence between points on the endocardial surface to the nearest registered EAM point. An empirically picked voltage was used as the threshold to identify target region in EAM. Based on the point-to-point correspondence, the target region defined on EAM was projected onto the endocardial contour in the MR image, which was fused to the planning CT for clinical target definition.

Results: The workflow was retrospectively tested on clinical data of a patient who received CRA. The target region derived from EAM was validated using clinical used target as the ground truth. The target volume derived using EAM data has a strong correlation with the clinical used PTV (>85% overlap). The process can be fully automated using Matlab and ESAPI scripts.

Conclusion: We have implemented and tested CPD to register EAM map data for CRA planning. The approach can utilize EAM information for CRA planning to promote optimal targeting in TPS.


Image-guided Surgery, MRI, Stereotactic Radiosurgery


IM/TH- Image-guided Surgery: Registration and Localization

Contact Email