The accuracy and reliability of 3D CT/MRI co-registration in planning epilepsy surgery

Abstract

Objective: To investigate the accuracy and reliability of 3D CT/MRI co-registration technique for the localization of implanted subdural electrodes in the routine epilepsy presurgical evaluation, in so doing assess its usefulness in planning the tailored resection of epileptic focus. Methods: Four external anatomic fiducial makers were used for co-registration of volumetric pre-implant brain MRI and post-implant head CT using Curry 5.0 software in 19 epilepsy presurgical candidates. The location of subdural electrodes derived from the co-registration was compared to that obtained by intra-operative digital photographs by using gyral/sulcal patterns and cortical vasculature as anatomic markers. Results: The mean localization error was 4.3 ± 2.5 mm in all 19 patients. However, the mean localization error was 3.1 ± 1.3 mm in 13 patients with all four reliable fiducial markers; whereas the mean localization error was 6.8 ± 2.4 mm in 6 patients with two or three reliable fiducial markers. Conclusion: Visualization of subdural electrode positions on a patient's cortex can be accurately performed in the routine clinical setting by 3D CT/MRI co-registration. However, the accuracy of co-registration is dependent upon having reliable surface fiducial markers. In practice, confirmation of location accuracy, such as with intra-operative digital photographs, is necessary for planning of tailored resective surgery. Significance: The combination of 3D CT/MRI co-registration and intra-operative digital photography techniques provides a practical and effective algorithm for the localization and validation of implanted subdural electrodes. © 2009 International Federation of Clinical Neurophysiology.