Methodology, outcome, safety and in vivo accuracy in traditional frame-based stereoelectroencephalography

Abstract

Background: Stereoelectroencephalography (SEEG) is an established diagnostic technique for the localization of the epileptogenic zone in drug-resistant epilepsy. In vivo accuracy of SEEG electrode positioning is of paramount importance since higher accuracy may lead to more precise resective surgery, better seizure outcome and reduction of complications.

Objective: To describe experiences with the SEEG technique in our comprehensive epilepsy center, to illustrate surgical methodology, to evaluate in vivo application accuracy and to consider the diagnostic yield of SEEG implantations.

Methods: All patients who underwent SEEG implantations between September 2008 and April 2016 were analyzed. Planned electrode trajectories were compared with post-implantation trajectories after fusion of pre- and postoperative imaging. Quantitative analysis of deviation using Euclidean distance and directional errors was performed. Explanatory variables for electrode accuracy were analyzed using linear regression modeling. The surgical methodology, procedure-related complications and diagnostic yield were reported.

Results: Seventy-six implantations were performed in 71 patients, and a total of 902 electrodes were implanted. Median entry and target point deviations were 1.54 mm and 2.93 mm. Several factors that predicted entry and target point accuracy were identified. The rate of major complications was 2.6%. SEEG led to surgical therapy of various modalities in 53 patients (69.7%).

Conclusions: This study demonstrated that entry and target point localization errors can be predicted by linear regression models, which can aid in identification of high-risk electrode trajectories and further enhancement of accuracy. SEEG is a reliable technique, as demonstrated by the high accuracy of conventional frame-based implantation methodology and the good diagnostic yield.

Keywords: Complications; Epilepsy surgery; In vivo accuracy; Stereoelectroencephalography; Stereotaxy.

Fig. 1

Methodology of electrode implantations. A and B: Planning of electrode trajectories using navigation software. C: Coordinates of the Leksell frame are set by the operating neurosurgeon. D: Introduction of the stylet through the guiding screw to a premeasured length to create the electrode tract. E: Insertion of the depth electrode through the screw. F: Aspect at the end of the procedure, after implantation of 13 depth electrodes and placement of fixation bolts

Fig. 2

Three cases of in vivo application accuracy measurements on postoperative CT scans. The planned trajectories are shown as solid lines. For visualization purposes, the CT bone window setting was used (−200 to 800 HU). A: Coronal and (B) axial reconstructions of the same electrode, showing optimal positioning of the implanted electrode in comparison with the planned trajectory. Target point localization error (TPLE) was 0.83 mm for this electrode. C: Minor deviation in the coronal plane of an orthogonal electrode after insertion in the skull. The TPLE was 2.70 mm. D: Major deformation of the electrode in the coronal plane, with evidence of deviation in the other planes as well, resulting in a TPLE of 9.03 mm. TPLEs were measured in three different planes and calculated using the Euclidean distance

Fig. 3

A: The concept of Euclidean distance. The tip of the planned trajectory is represented by point P, and the tip of the actual electrode is represented by point Q. The arrow is the Euclidean distance between both points. B: The Euclidean distance formula. For two points, the coordinates x, y and z are determined, and the Euclidean distance is defined as the square root of the sum of the squares of the difference between these coordinates

Fig. 4

Density scatterplots of electrode target directional errors. Higher density areas represent more electrodes with the same directional errors. In the left pane, the directional errors in the medial-lateral X-direction (horizontal axis) and the anterior-posterior Y-direction (vertical axis) are shown. The right side plot shows directional errors in the medial-lateral X-direction (horizontal axis) and the caudal-cranial Z-direction (vertical axis). The graphs illustrate a small deviation in the lateral and cranial directions