In Vivo Evaluation of Image Registration Techniques During Endovascular Repair

Overview

Fluoroscopy is traditionally used for guidance of endovascular aortic repair (EVAR). In order to minimize exposure to radiation and nephrotoxic contrast medium, it is possible to generate a navigation road map by registering the intraoperative images with preoperative computed tomography angiograph (CTA). In modern hybrid operating rooms, several commercial solutions for this exists today (e.g. Siemens Syngo iPilot, Phillips Vesselnavigator). In order to register (fuse) the preoperative and intraoperative imaging to each other, a registration algorithm has to be applied. Sufficient accuracy of this algorithm is crucial, for if it fails the road map cannot be used for intraoperative navigation. The purpose of this study is therefore to evaluate two different 3D registration algorithms with regard to registration accuracy.

Two different registration algorithms will be tested: 1. registration algorithm - centerline registration algorithm (Sintef, Trondheim) A catheter equipped with a position sensor records the travelled path of the instrument, which is assumed to be close to the centerline of the aorta. From the preoperative computer tomography scan (CT) the aorta is segmented and vessel centerline is extracted. The centerline from the position sensor recordings is registered to the extracted centerline of the preoperative CT. 2. 3D-3D registration algorithm (Syngo iPilot, Siemens Medical Solutions, Munich). Intraoperative cone-beam-CT (CBCT) is registered to the preoperative CT

Conditions

Interventions