Stereotactic registration using cone-beam computed tomography

Abstract

Objective: Stereotactic registration for deep brain stimulation surgery is often performed after halo placement on an MRI or CT with fiducials. Cone-beam computed tomography scanners (CBCT) have been used for lead confirmation in DBS surgery. A new CBCT scanner has an enlarged field of view now allowing for visualization of fiducials, and the potential for intraoperative registration. The aim of this retrospective study was to evaluate the accuracy and precision of stereotactic registration using images obtained in intraoperative CBCT.

Patients and methods: The registration accuracy and precision of CBCT (O-arm O2, Medtronic, Dublin, Ireland) was compared to CT registration study obtained in the same patient (n = 10). Stereotactic coordinate differences were compared. In a second analysis the end-to-end accuracy and precision between the surgical target and lead position was analyzed on postoperative MRI in cases where CBCT was used for registration (n = 31 leads).

Results: The average radial distance of the stereotactic coordinate using the CBCT and CT registration studies was not clinically different (0.46 ± 0.17 mm (max 0.81)). The registration image maximum rod error was more accurate for CBCT (0.50 ± 0.12 mm (max 0.7)), than the CT registration study (1.02 ± 0.63 mm) (max 1.8) (P = 0.018). On average 26 min was saved using only the CBCT to perform registration (P < 0.001). The radial error in the end-to-end analysis was 1.07 ± 0.67 (max 2.4) measured on postoperative MRI.

Conclusion: Registration using a CBCT image is accurate, and using this workflow yields accurate and precise DBS lead placement.

Keywords: Cone-Beam computed tomography; Deep brain stimulation; Stereotactic techniques; Surgical accuracy.