The Effect of Vascular Segmentation Methods on Stereotactic Trajectory Planning for Drug-Resistant Focal Epilepsy: A Retrospective Cohort Study

Computer-assisted stereoelectroencephalography planning: center-specific priors enhance planning

Objectives

This study aims to refine computer-assisted planning (CAP) of SEEG implantations by adding spatial constraints from prior SEEG trajectories ("Priors") to improve safety and reduce manual adjustments, without increasing planning time.

Methods

Retrospective validation based on 159 previously implanted trajectories (11 cases) planned by the clinical standard CAP and CAP constrained with spatial priors ("CAP + Priors"). Constraints included 31 target and 51 entry zones, created from 98 consecutive patients (763 implanted SEEG trajectories). Each of the 159 previously implanted trajectories was planned by two fellows, once with CAP and once with CAP + Priors, in a randomized order. The time taken to generate the initial computer-generated plan (T1) and the user-edited final plan (T2) were recorded together with the proportions of electrodes that required subsequent adjustments. Clinical implantability was assessed via a blinded review of each trajectory by five independent epilepsy neurosurgeons with expertise in SEEG implantation.

Results

Expert raters considered 88.5% of trajectories implantable, with no difference in acceptability between CAP alone and CAP + Priors (p = 0.79). Median (IQR) T1 for CAP to produce complete automated implantation was 4.6 (0.85) min vs. CAP + Priors was 6.3 (2.6) min (p = 0.03). There was no significant difference in T2 (time to complete surgeon-edited plan): CAP median (IQR) 105 (22) min, and CAP + Priors median (IQR) 96 (68) min (p = 0.92). The CAP + Priors risk score was significantly lower than that for the previously actually implanted trajectories for the 11 plans analyzed (p = 0.004), and no different from CAP alone planning. A significant reduction was observed in manual adjustments required with CAP + Priors in the cingulate gyrus.

Conclusion

Using spatial priors from previous implantations enhances SEEG CAP and increases the granularity of trajectory planning. This approach facilitates more standardized planning and allows for the incorporation of experience from multiple expert centers, decreasing the risk of the resultant trajectories and reducing the proportion of trajectories that require manual planning without significantly increasing planning time.

The Value of SINO Robot and Angio Render Technology for Stereoelectroencephalography Electrode Implantation in Drug-Resistant Epilepsy

BACKGROUND

 Stereoelectroencephalography (SEEG) electrodes are implanted using a variety of stereotactic technologies to treat refractory epilepsy. The value of the SINO robot for SEEG electrode implantation is not yet defined. The aim of the current study was to assess the value of the SINO robot in conjunction with Angio Render technology for SEEG electrode implantation and to assess its efficacy.

METHODS

 Between June 2018 and October 2020, 58 patients underwent SEEG electrode implantation to resect or ablate their epileptogenic zone (EZ). The SINO robot and the Angio Render technology was used to guide the electrodes and visualize the individual vasculature in a three-dimensional (3D) fashion. The 3D view functionality was used to increase the safety and accuracy of the electrode implantation, and for reducing the risk of hemorrhage by avoiding blood vessels.

RESULTS

 In this study, 634 SEEG electrodes were implanted in 58 patients, with a mean of 10.92 (range: 5-18) leads per patient. The mean entry point localization error (EPLE) was 0.94 ± 0.23 mm (range: 0.39-1.63 mm), and the mean target point localization error (TPLE) was 1.49 ± 0.37 mm (range: 0.80-2.78 mm). The mean operating time per lead (MOTPL) was 6. 18 ± 1.80 minutes (range: 3.02-14.61 minutes). The mean depth of electrodes was 56.96 ± 3.62 mm (range: 27.23-124.85 mm). At a follow-up of at least 1 year, in total, 81.57% (47/58) patients achieved an Engel class I seizure freedom. There were two patients with asymptomatic intracerebral hematomas following SEEG electrode placement, with no late complications or mortality in this cohort.

CONCLUSIONS

 The SINO robot in conjunction with Angio Render technology-in SEEG electrode implantation is safe and accurate in mitigating the risk of intracranial hemorrhage in patients suffering from drug-resistant epilepsy.

Use of Compressed Sensing Accelerated, Low-Velocity Encoded, Isotropic Resolution, Phase Contrast Magnetic Resonance Angiography for SEEG Electrode Implantation

OBJECTIVE

We assessed the feasibility of using compressed sensing accelerated, low-velocity encoded, isotropic resolution phase contrast (CLIP) magnetic resonance angiography (MRA) for avascular trajectory planning of stereoelectroencephalography.

METHODS

Ten healthy subjects (1 woman and 9 men; age, 33.6 ± 9.0 years) and 20 consecutive patients (12 female patients; age, 22 ± 13.6 years) were enrolled in the present study. The healthy subjects underwent CLIP-MRA, and 3 other phase contrast MRA protocols with conventional parallel imaging (PI) acceleration, including low resolution with twofold PI (PI2), high resolution (HR) with fivefold PI (PI5), and HR-PI2. The patients underwent CLIP-MRA and computed tomography angiography (CTA). The image qualities were evaluated. The numbers and locations of trajectory-vessel conflict detected using CLIP-MRA were noted.

RESULTS

With similar scan durations, CLIP-MRA achieved higher spatial resolution compared with low resolution with PI2 and detected significantly more branches compared with HR-PI5. With the same spatial resolution, the signal/noise and contrast/noise ratios of CLIP-MRA were higher than those with HR-PI2 with a shorter scan duration. For the 12 adult patients (10 female patients; 28.8 ± 12.7 years), CLIP-MRA had better signal/noise and contrast/noise ratios than CTA. The trajectory had required modification for 14 of the 20 patients (70%), with a proportion of trajectory modification of 10.7% (23 of 215 electrodes). The middle meningeal artery, cortical vessel, and skull vessel were the main vessels with conflict (n = 11, n = 7, and n = 5, respectively).

CONCLUSIONS

In the present study, CLIP-MRA provided a clear cortical vascular display noninvasively without intravascular contrast and radiation. The middle meningeal artery and diploic and emissary veins were the main conflict vessels and could be clearly displayed using CLIP-MRA but not CTA.

Validation of SART 3.5D algorithm for cerebrovascular dynamics and artery versus vein classification in presurgical 3D digital subtraction angiographies

Classification of arteries and veins in cerebral angiograms can increase the safety of neurosurgical procedures, such as StereoElectroEncephaloGraphy, and aid the diagnosis of vascular pathologies, as arterovenous malformations. We propose a new method for vessel classification using the contrast medium dynamics in rotational digital subtraction angiography (DSA). After 3D DSA and angiogram segmentation, contrast enhanced projections are processed to suppress soft tissue and bone structures attenuation effect and further enhance the CM flow. For each voxel labelled as vessel, a time intensity curve (TIC) is obtained as a linear combination of temporal basis functions whose weights are addressed by simultaneous algebraic reconstruction technique (SART 3.5D), expanded to include dynamics. Each TIC is classified by comparing the areas under the curve in the arterial and venous phases. Clustering is applied to optimize the classification thresholds. On a dataset of 60 patients, a median value of sensitivity (90%), specificity (91%), and accuracy (92%) were obtained with respect to annotated arterial and venous voxels up to branching order 4–5. Qualitative results are also presented about CM arrival time mapping and its distribution in arteries and veins respectively. In conclusion, this study shows a valuable impact, at no protocol extra-cost or invasiveness, concerning surgical planning related to the enhancement of arteries as major organs at risk. Also, it opens a new scope on the pathophysiology of cerebrovascular dynamics and its anatomical relationships.

On the Use of Digital Subtraction Angiography in Stereoelectroencephalography Surgical Planning to Prevent Collisions with Vessels

OBJECTIVE

Stereoelectroencephalography (SEEG) consists of the implantation of microelectrodes for the electrophysiological characterization of epileptogenic networks. To reduce a possible risk of intracranial bleeding by vessel rupture during the electrode implantation, the stereotactic trajectories must follow avascular corridors. The use of digital subtraction angiography (DSA) for vascular visualization during planning is controversial due to the additional risk related to this procedure. Here we evaluate the utility of this technique for planning when the neurosurgeon has it available together with gadolinium-enhanced T1-weighted magnetic resonance sequence (T1-Gd) and computed tomography angiography (CTA).

METHODS

Twenty-two implantation plans for SEEG were initially done using T1-Gd imaging (251 trajectories). DSA was only used later during the revision process. In 6 patients CTA was available at this point as well. We quantified the position of the closest vessel to the trajectory in each of the imaging modalities.

RESULTS

Two thirds of the trajectories that appeared vessel free in the T1-Gd or CTA presented vessels in their proximity, as shown by DSA. Those modifications only required small shifts of both the entry and target point, so the diagnostic aims were preserved.

CONCLUSIONS

T1-Gd and CTA, despite being the most commonly used techniques for SEEG planning, frequently fail to reveal vessels that are dangerously close to the trajectories. Higher-resolution vascular imaging techniques, such as DSA, can provide the neurosurgeon with crucial information about vascular anatomy, resulting in safer plans.

Refining Planning for Stereoelectroencephalography: A Prospective Validation of Spatial Priors for Computer-Assisted Planning With Application of Dynamic Learning

Objective: Stereoelectroencephalography (SEEG) is a procedure in which many electrodes are stereotactically implanted within different regions of the brain to estimate the epileptogenic zone in patients with drug-refractory focal epilepsy. Computer-assisted planning (CAP) improves risk scores, gray matter sampling, orthogonal drilling angles to the skull and intracerebral length in a fraction of the time required for manual planning. Due to differences in planning practices, such algorithms may not be generalizable between institutions. We provide a prospective validation of clinically feasible trajectories using “spatial priors” derived from previous implantations and implement a machine learning classifier to adapt to evolving planning practices.

Methods: Thirty-two patients underwent consecutive SEEG implantations utilizing computer-assisted planning over 2 years. Implanted electrodes from the first 12 patients (108 electrodes) were used as a training set from which entry and target point spatial priors were generated. CAP was then prospectively performed using the spatial priors in a further test set of 20 patients (210 electrodes). A K-nearest neighbor (K-NN) machine learning classifier was implemented as an adaptive learning method to modify the spatial priors dynamically.

Results: All of the 318 prospective computer-assisted planned electrodes were implanted without complication. Spatial priors developed from the training set generated clinically feasible trajectories in 79% of the test set. The remaining 21% required entry or target points outside of the spatial priors. The K-NN classifier was able to dynamically model real-time changes in the spatial priors in order to adapt to the evolving planning requirements.

Conclusions: We provide spatial priors for common SEEG trajectories that prospectively integrate clinically feasible trajectory planning practices from previous SEEG implantations. This allows institutional SEEG experience to be incorporated and used to guide future implantations. The deployment of a K-NN classifier may improve the generalisability of the algorithm by dynamically modifying the spatial priors in real-time as further implantations are performed.