Frame-based and robot-assisted insular stereo-electroencephalography via an anterior or posterior oblique approach

There is an increasing interest in stereo-electroencephalography (SEEG) for invasive evaluation of insular epilepsy. The implantation of insular SEEG electrodes, however, is still challenging due to the anatomical location and complex functional segmentation in both an anteroposterior and ventrodorsal (i.e., superoinferior) direction. While the orthogonal approach (OA) is the shortest trajectory to the insula, it might insufficiently cover these networks. In contrast, the anterior approach (AOA) or posterior oblique approach (POA) has the potential for full insular coverage, with fewer electrodes bearing a risk of being more inaccurate due to the longer trajectory. Here, the authors evaluated the implantation accuracy and the detection of epilepsy-related SEEG activity with AOA and POA insular trajectories.

This retrospective study evaluated the accuracy of 220 SEEG electrodes in 27 patients. Twelve patients underwent a stereotactic frame-based procedure (frame group), and 15 patients underwent a frameless robot-assisted surgery (robot group). In total, 55 insular electrodes were implanted using the AOA or POA considering the insular anteroposterior and ventrodorsal functional organization. The entry point error (EPE) and target point error (TPE) were related to the implantation technique (frame vs robot), the length of the trajectory, and the location of the target (insular vs noninsular). Finally, the spatial distribution of epilepsy-related SEEG activity within the insula is described.

There were no significant differences in EPE (mean 0.9 ± 0.6 for the nonsinsular electrodes and 1.1 ± 0.7 mm for the insular electrodes) and TPE (1.5 ± 0.8 and 1.6 ± 0.9 mm, respectively), although the length of trajectories differed significantly (34.1 ± 10.9 and 70.1 ± 9.0 mm, repsectively). There was a significantly larger EPE in the frame group than in the robot group (1.5 ± 0.6 vs 0.7 ± 0.5 mm). However, there was no group difference in the TPE (1.5 ± 0.8 vs 1.6 ± 0.8 mm). Epilepsy-related SEEG activity was detected in 42% (23/55) of the insular electrodes. Spatial distribution of this activity showed a clustering in both anteroposterior and ventrodorsal directions. In purely insular onset cases, subsequent insular lesionectomy resulted in a good seizure outcome.

The implantation of insular electrodes via the AOA or POA is safe and efficient for SEEG implantation covering both anteroposterior and ventrodorsal functional organization with few electrodes. In this series, there was no decrease in accuracy due to the longer trajectory of insular SEEG electrodes in comparison with noninsular SEEG electrodes. The results of frame-based and robot-assisted implantations were comparable.