Computer-assisted planning for the insertion of stereoelectroencephalography electrodes for the investigation of drug-resistant focal epilepsy: an external validation study.
Vejay N Vakharia,Rachel Sparks,Roman Rodionov,Sjoerd B Vos,Christian Dorfer,Jonathan Miller,Daniel Nilsson,Martin Tisdall,Stefan Wolfsberger,Andrew W McEvoy,Anna Miserocchi,Gavin P Winston,Aidan G O'Keeffe,Sebastien Ourselin,John S Duncan
DOI: https://doi.org/10.3171/2017.10.JNS171826
IF: 5.408
2019-10-03
Journal of Neurosurgery
Abstract:OBJECTIVE One-third of cases of focal epilepsy are drug refractory, and surgery might provide a cure. Seizure-free outcome after surgery depends on the correct identification and resection of the epileptogenic zone. In patients with no visible abnormality on MRI, or in cases in which presurgical evaluation yields discordant data, invasive stereoelectroencephalography (SEEG) recordings might be necessary. SEEG is a procedure in which multiple electrodes are placed stereotactically in key targets within the brain to record interictal and ictal electrophysiological activity. Correlating this activity with seizure semiology enables identification of the seizure-onset zone and key structures within the ictal network. The main risk related to electrode placement is hemorrhage, which occurs in 1% of patients who undergo the procedure. Planning safe electrode placement for SEEG requires meticulous adherence to the following: 1) maximize the distance from cerebral vasculature, 2) avoid crossing sulcal pial boundaries (sulci), 3) maximize gray matter sampling, 4) minimize electrode length, 5) drill at an angle orthogonal to the skull, and 6) avoid critical neurological structures. The authors provide a validation of surgical strategizing and planning with EpiNav, a multimodal platform that enables automated computer-assisted planning (CAP) for electrode placement with user-defined regions of interest. METHODS Thirteen consecutive patients who underwent implantation of a total 116 electrodes over a 15-month period were studied retrospectively. Models of the cortex, gray matter, and sulci were generated from patient-specific whole-brain parcellation, and vascular segmentation was performed on the basis of preoperative MR venography. Then, the multidisciplinary implantation strategy and precise trajectory planning were reconstructed using CAP and compared with the implemented manually determined plans. Paired results for safety metric comparisons were available for 104 electrodes. External validity of the suitability and safety of electrode entry points, trajectories, and target-point feasibility was sought from 5 independent, blinded experts from outside institutions. RESULTS CAP-generated electrode trajectories resulted in a statistically significant improvement in electrode length, drilling angle, gray matter-sampling ratio, minimum distance from segmented vasculature, and risk (p < 0.05). The blinded external raters had various opinions of trajectory feasibility that were not statistically significant, and they considered a mean of 69.4% of manually determined trajectories and 62.2% of CAP-generated trajectories feasible; 19.4% of the CAP-generated electrode-placement plans were deemed feasible when the manually determined plans were not, whereas 26.5% of the manually determined electrode-placement plans were rated feasible when CAP-determined plans were not (no significant difference). CONCLUSIONS CAP generates clinically feasible electrode-placement plans and results in statistically improved safety metrics. CAP is a useful tool for automating the placement of electrodes for SEEG; however, it requires the operating surgeon to review the results before implantation, because only 62% of electrode-placement plans were rated feasible, compared with 69% of the manually determined placement plans, mainly because of proximity of the electrodes to unsegmented vasculature. Improved vascular segmentation and sulcal modeling could lead to further improvements in the feasibility of CAP-generated trajectories.
surgery,clinical neurology