Robotic-Guided Stereoelectroencephalography for Invasive Epilepsy Monitoring

[Narrator 1] Welcome to our video on robotic-guided stereoelectroencephalography for invasive epilepsy monitoring. Stereo EEG, or SEEG, was pioneered by Dr. Jean Bancaud and Talairach in the 1960s in Paris. This method refers to the stereotactic implantation of intracerebral electrodes according to an anatomo-electro-clinical methodology, which gave rise to the concept of sampling and recording from networks in the brain during epileptic seizures. This gave a three-dimensional perspective and now underscores our ability to sample spatiotemporal dynamical evolution in seizure events through intracerebral EEG electrodes. With superposition of vascular imaging, together with the Talairach grid, as illustrated by the Szikla Atlas, it becomes clear and imperative the careful stereotactic methods are important to the safe implantation of electrodes into intended targets along trajectories that minimize injury to both superficial and deep vessels. A few words about SEEG planning. Here at the Cleveland Clinic, we adhere to the French School of Epileptology as per the Talairach and Bancaud era. Apostrophes in our implantation maps denote the left side, for example, A prime as opposed to A. We conventionally perform implantations using orthogonal electrodes. We would like to highlight that the trajectory is as important as the intended targets to better understand network dynamics. Also, particular regions of interest. For example, the insula, or the posterior orbital frontal region, may require oblique trajectories for implantation. And finally, it should be noted that risky trajectories, for example, the insula, can still be performed by surgeons who are specifically trained in the method typically at experienced epilepsy centers. Notably, the number of electrodes used must sample the anatomo-electro-clinical hypothesis that is formulated by a comprehensive multidisciplinary epilepsy team. And so as an example, this may range from 12 to 20 electrodes or so. Here, we offer a planning example with a pre-implantation map, showing the exploration of mesial temporal and anterior pericerebral networks favoring orthogonal trajectories using 13 electrodes. In the table, we have a standard nomenclature for each electrodes, with the apostrophe denoting the left side. There are targets and entry sites that are defined here, but again, it is important to highlight the trajectory planning itself, which is conventionally done on a robotic system.

[Narrator 2] Here, we have a presentation for one of the cases used for filming. Ethical considerations for this case presentation have been addressed by our institution's internal review board. The patient is a 46-year-old with a history of drug-resistant epilepsy and the setting of left hemispheric cortical dysplasia associated with polymicrogyria. He had extensive pre-surgical workup, which resulted in a pre-SEEG hypothesis of focal epilepsy, despite the large hemispheric malformation. This hypothesis is the basis of the SEEG implantation plan, which is constructed using the principles discussed in the aforementioned planning example. The final trajectories are planned using the robotic interface and planning software.

[Narrator 1] Here's an example of the robotic interface and the planning software used in preparing for the SEEG procedure. Intracerebral electrodes are planned sequentially, as this one shown here in the head of the hippocampus. We define a two-millimeter diameter that is applied to each electrode as our error cut off. The electrode is designated as B prime. We select the target in the head of the hippocampus itself and also think about the entry point so that we preserve an orthogonal electrode and trajectory. Of course, vessels at the surface are identified, and we work around them defining an entry point. As we move to the cross-sectional view, we are able to visualize for any vessel collision, as shown here. Notably, we have fused CT and MR sequences together. Our scanning protocols include double-contrasted and vessel imaging, including MRA, to facilitate visualization of arteries, as well as the venous anatomy. Here, you see proximity to the superior temporal sulcus, again, pointing out the importance of the trajectory. At the surface, we are in close proximity here to the venous anatomy, and adjustments are made accordingly to avoid the superficial vessels. The trajectory is then reassessed and we convert to the CT scan windows to evaluate once more. The CT is useful, particularly to identify some of the venous structures at the surface, as well as to avoid any air cells that could pose a risk for CSF leakage. Now, fast forwarding ahead, we have completed the plan, in this case with multiple electrodes spaced out. Here we see, for example, Q, R prime, R prime, and S prime, through the frontal parietal operculum and ending in the insula. And great care is taken to evaluate each and every single electrode along its trajectory length to ensure for no collisions with vessels or other structures. We want to maximize and capitalize on the sampling of gray matter as well, including the depths of sulci. And finally, as you can see, some electrodes are deeper than others, but we do have a collection of the electrodes now showing all on the same space. And with rotation of the head In the 3D view, we can actually inspect from the inside to ensure that there is no collision identified, either notably between the oblique electrode trajectory and the other orthogonally-placed electrodes as shown here.

[Narrator 2] A final check is performed by the operative team on the day of surgery. The monitoring team takes measurements off of the robotic system to select electrodes with the appropriate length and contact spacing. Next, the Leksell Frame is attached to the patient's head. The Leksell Frame is a good option for head fixation, as it allows access to the face for facial recognition and to both sides of the head for bilateral electrode placement. There are various options for registering the robot to the patient's anatomy, including facial recognition, and in this case, using non-invasive landmarks rigidly attached to the Leksell Frame. An intraoperative O-arm spin is obtained with the frame attached to assist in registration for this method. The use of rigidly-attached fiducials with an intraoperative O-arm spin provides the highest accuracy and reliability and fits well within a workflow using the robotic arm. Once registration is complete, the robot is attached to the head frame. Next, the surgical field is prepped and sterile drapes are placed. The robot base is draped, and subsequently, the sterile attachment to the robot arm is attached over top of the drape. The robot arm is directed into position on the trajectory for each electrode. The axial movements of the arm allow for precise movement along the electrode trajectory. The burr hole is drilled through the guard on the robot arm. The guard on the drill acts as a stop, preventing the surgeon from plunging through the bone. The axial movements of the robot arm can be used to finally adjust the depth of drilling, thus minimizing time-consuming manipulations of the guard. A dural opening is made by applying monopolar cautery through a probe. The bolt, which secures each electrode, is screwed in place through the robot arm. The electrode securement cap is then screwed onto each bolt. The electrode is then advanced through the guide on the robot arm, along its trajectory to the premeasured appropriate depth. It is then secured at that depth with the cap. Depending on the electrode design, it may be necessary to place the electrode freehand, rather than through the robot arm in order to free the tail of the electrode. It is secured at its exit from the bolt and the axial movement of the robot arm is used to create space. Subsequently, the tail of the electrode is pulled through the guide on the robot arm. Occasionally, trace bleeding at the site of a bolt is encountered. This can usually be addressed with gentle irrigation. Copious bleeding should be investigated further. Each electrode is labeled and recorded to aid in later interpretation. Once all the electrodes have been placed, grounds are placed and secured, the entire setup is then connected to recording equipment and an intraoperative recording is obtained. This is interpreted by the neurology and neurosurgery teams in order to ensure adequate signals are being obtained from every electrode. Finally, the robot is disconnected from the frame. All of the wires are packaged in a plastic bag. The electrode sites are padded and wrapped in sterile fashion. The head frame is removed and the head is wrapped. A representative example of electrode placement is illustrated in this postoperative X-ray, which demonstrates that the electrodes are grossly in the correct position without any visible bends or kinks. To examine the efficiency of the method, we computed the surgical time, as well as the total time in room for SEEG patients over the course of a year. We found that the average time in room per electrode was about 15 minutes, but that the efficiency improves with increasing numbers of electrodes placed to as low as 9.4 minutes per electrode. This likely reflects that the setup and takedown time of the robot are constant, while the improved efficiency of the robotic technique is compounded as the number of trajectories increases. In conclusion, the use of stereotactic frame-based methods for SEEG placement is well-established, safe, and precise. The use of robotic guidance may help to increase the speed and ease of electrode placement, while maintaining the accuracy and precision of these classical techniques. And we have found in particular that the efficiency benefits of the robotic technique are best appreciated when a larger number of electrodes are placed. We hope you have enjoyed our presentation on robotic-guided stereoelectroencephalography.