Computational analysis of application accuracy in stereoelectroencephalography in over 2500 consecutive trajectories
Abstract number :
1.403
Submission category :
9. Surgery / 9A. Adult
Year :
2025
Submission ID :
1176
Source :
www.aesnet.org
Presentation date :
12/6/2025 12:00:00 AM
Published date :
Authors :
Greydon Gilmore, PhD – Schulich School of Medicine and Dentistry, Western University
Presenting Author: Amit Persad, MD – Western University
Arun Thurairajah, MSc – Schulich School of Medicine and Dentistry, Western University
Alaa Taha, PhD – Western University
Mohamad Abbass, MD – Western University
Brendan Santyr, MD – Western University
Khalid Alorabi, MD – Schulich School of Medicine and Dentistry, Western University
Amit Persad, MD – Western University
Andrew Parrent, MD – Western University
Keith W MacDougall, MD – Western University
David A Steven, MD, MPH, FRCSC, FACS – Western University
Jonathan C Lau, MD, PhD – Western University
Rationale: Stereoelectroencephalography (SEEG) is a diagnostic procedure involving surgical implantation of electrodes in which utility depends on accurate targeting. Error in SEEG application can be measured in Euclidean distance, radial distance, and angle error. Here, we present our consecutive series of SEEG implantation with comprehensive description of implantation accuracy using these metrics. Furthermore, we divide the electrodes into standard trajectories and compare their accuracies by class.
Methods: A consecutive series of patients between 2017 and 2024 who underwent robotic SEEG implantation were included in this retrospective review. Baseline demographics were recorded, as well as electrode coverage, overall case time (including anaesthesia time), operation time, radiation exposure, and complications. Application accuracy was retrospectively computed for each electrode in a consecutive series of SEEG implantation (frame-based and robot-assisted). Specifically, Euclidean and radial distance were calculated between the planned target point (based on the preoperative planning MRI) and the tip of the actual electrode (measured using postoperative CT).
Results: Data from 2658 electrode implantations were analysed. There were more electrodes placed during robot-assisted application (11.94 vs 10.23 electrodes, p=0.02). Mean target and entry point Euclidean errors were 2.27 ± 1.19 mm and 1.37 ± 0.99 mm for robot-assisted SEEG and 3.08 ± 1.83 mm and 2.30 ± 1.71 mm for frame-based SEEG (p< 0.001). Mean target and entry point radial errors were 1.58 ± 1.03 mm and 1.21 ± 0.84 mm for robot-assisted, and 2.26 ± 1.57 mm and 1.99 ± 1.48 mm for frame-based (p< 0.001). The least accurately targeted structures were the temporal pole and posterior insula with radial errors of 2.66 mm (1.80-3.52) and 2.05 mm (1.80-2.30) respectively. The most accurately targeted structures were the posterior and anterior supplementary somatosensory motor areas with a radial error of 0.86 mm (0.63-1.09) and 1.00 mm (0.73-1.26) respectively.
Surgery