At the London Health Sciences Centre Epilepsy Program, stereotactically implanted depth electrodes have largely replaced subdural electrodes in the presurgical investigation of patients with drug-resistant epilepsy over the past 4 years. The rationale for this paradigm shift was more experience with, and improved surgical techniques for, stereoelectroencephalography, a possible lower-risk profile for depth electrodes, better patient tolerability, shorter operative time, as well as increased recognition of potential surgical targets that are not accessible to subdural electrodes.

Background: Stereoelectroencephalography has been in regular use at the Montreal Neurological Institute since 1972. The technique has been in constant evolution to incorporate advances in materials, imaging, and robotics technology. MRI-compatible electrodes were introduced in 2007 and robotics in 2011. Here we report on the technique, safety, and advantages of our current method of stereoelectroencephalography implantation. Methods: We retrospectively reviewed all patients who underwent stereoelectroencephalography by the senior author. Technical, clinical, and radiological complications, and postimplantation outcomes were analyzed. Only patients implanted with MRI-compatible electrodes were included to review MRI abnormalities with electrodes in situ. Results: A total of 53 patients were implanted with 550 electrodes (average=10.4 per patient), for an average duration of 14.6 days. There was no mortality, infection, or new neurologic deficit. Two patients had a superficial screw plunge without clinical consequence. Four patients demonstrated asymptomatic MRI abnormalities (7.54% per patient, or 0.72% per electrode). MRI with electrodes in situ was used for neuronavigation in all 29 who underwent resection and yielded a histopathological diagnosis of focal cortical dysplasia in 15 MRI-negative patients. Conclusions: The technique of stereoelectroencephalography described here was associated with no clinical morbidity although not without technical complications or radiologic (MRI) abnormalities. We should therefore remain vigilant in refining the technique and minimizing the number of electrodes required to answer a well-developed hypothesis regarding the epileptogenic zone. The use of MRI-compatible electrodes allowed neuronavigation using the images with the electrodes in situ, which was useful to tailor the eventual definitive resection and in localizing MRI-negative lesions.