Background: The aim of our study was to evaluate discrepancies between the electrophysiologically and MRI-defined subthalamic nucleus (STN) in order to contribute to the ongoing debate of whether or not microelectrode recording (MER) provides additional information to image-guided targeting in deep brain stimulation.
Methods: Forty-four STNs in 22 patients with Parkinson's disease were investigated. The three-dimensional MRI-defined STN was derived from segmentations of axial and coronal T2-weighted images. The electrophysiological STNs were generated from intraoperative MERs in 1,487 locations. The stereotactical coordinates of positive and negative STN recordings were re-imported to the planning software, where a three-dimensional reconstruction of the electrophysiological STN was performed and fused to the MRI data set. The estimated borders of the MRI- and MER-STN were compared. For statistical analysis Student's t, Mann-Whitney rank sum and Fisher's exact tests were used.
Results: MER-STN volumes, which were found outside the MRI-STN, ranged from 0 mm(3) to 87 mm(3) (mean: 45 mm(3)). A mean of 44% of the MER-STN volumes exceeded the MRI-STN (maximum: 85.1%; minimum: 15.1 %); 53.4% (n = 793) of the microelectrode recordings were concordant and 46.6% (n = 694) discordant with the MRI-defined anatomical STN. Regarding the dorsal borders, we found discrepancies between the MER- and MRI-STN of 0.27 mm (= mean; SD: 0.51 mm) on the first operated side and 1.51 mm (SD: 1.5 mm) on the second (p = 0.010, t-test).
Conclusions: MER provides additional information to high-resolution anatomical MR images and may help to detect the amount and direction of brain shift.