Accuracy of deep brain stimulation electrode placement using intraoperative computed tomography without microelectrode recording

Abstract

Object In this prospective study the authors' objective was to evaluate the accuracy of deep brain stimulation (DBS) electrode placement using image guidance for direct anatomical targeting with intraoperative CT. Methods Preoperative 3-T MR images were merged with intraoperative CT images for planning. Electrode targets were anatomical, based on the MR images. A skull-mounted NexFrame system was used for electrode placement, and all procedures were performed under general anesthesia. After electrode placement, intraoperative CT images were merged with trajectory planning images to calculate accuracy. Accuracy was assessed by both vector error and deviation off the planned trajectory. Results Sixty patients (33 with Parkinson disease, 26 with essential tremor, and 1 with dystonia) underwent the procedure. Patient's mean age was 64 ± 9.5 years. Over an 18-month period, 119 electrodes were placed (all bilateral, except one). Electrode implant locations were the ventral intermediate nucleus (VIM), globus pallidus internus (GPI), and subthalamic nucleus (STN) in 25, 23, and 12 patients, respectively. Target accuracy measurements were as follows: mean vector error 1.59 ± 1.11 mm and mean deviation off trajectory 1.24 ± 0.87 mm. There was no statistically significant difference between the accuracy of left and right brain electrodes. There was a statistically significant (negative) correlation between the distance of the closest approach of the electrode trajectory to the ventricular wall of the lateral ventricle and vector error (r2 = −0.339, p < 0.05, n = 76), and the deviation from the planned trajectory (r2 = −0.325, p < 0.05, n = 77). Furthermore, when the distance from the electrode trajectory and the ventricular wall was < 4 mm, the correlation of the ventricular distance to the deviation from the planned trajectory was stronger (r2 = −0.419, p = 0.05, n = 19). Electrodes placed in the GPI were significantly more accurate than those placed in the VIM (p < 0.05). Only 1 of 119 electrodes required intraoperative replacement due to a vector error > 3 mm. In this series there was one infection and no intraparenchymal hemorrhages. Conclusions Placement of DBS electrodes using an intraoperative CT scanner and the NexFrame achieves an accuracy that is at least comparable to other methods.