Applicative accuracy at entry and target points for robot-assisted depth electrodes implantation for drug-resistant epilepsy. Monocentric retrospective analysis of 1090 consecutive trajectories - Université de Picardie Jules Verne
Article Dans Une Revue Brain and Spine Année : 2023

Applicative accuracy at entry and target points for robot-assisted depth electrodes implantation for drug-resistant epilepsy. Monocentric retrospective analysis of 1090 consecutive trajectories

Résumé

Background: The implantation of SEEG electrodes with robotic assistance has become a standard procedure for invasive intracranial recordings in drug-resistant epilepsy. As compared with laser registration, the use of bony fiducials enables to reach a superior accuracy but requires the placement of 5 to 6 bony markers prior to imaging. Defining bony landmarks on Leskell G-frame is an option whose theoretical registration accuracy (rms value provided by the software) seems satisfactory. However data regarding the applicative accuracy on that method on a great number of procedures in real word conditions has been largely unknown until now. The objective of the present study was to assess the applicative accuracy obtained through this registration method. Methods: Monocentric retrospective analysis of the concordance between the planned and actual trajectories in 73 consecutive patients undergoing intracranial recordings for intractable epilepsy between Oct 2018 and Feb 2023. For each lead, the coordinates of the entry point and target point of the planned trajectory and those of the actual trajectory defined on the postoperative CT were automatically extracted via a dedicated MatLab® script. The Euclidian distance of error in mm at the entry point and target point were computed, as well as the angular deviation in degree between the trajectories. The statistical analysis (uni, multivariate, mixed model) was conducted with R software® (Version 2022.12.0+353) Results: A total of 1090 trajectories (73 patients) were evaluated. The mean Euclidian distance of error at the entry point was 0.86 mm (+/-0.58). At the target point, it was 2.15 mm (+/- 0.72). The mean angle of deviation in degree was 1.23. For orthogonal trajectories, the mean error was 0.55 and 0.43 mm in y and z respectively. The orthogonality of the trajectory was significantly associated with a smaller error at the entry point and target point (p<0.0001). No correlation was found between the rms value of registration quality and the error at the entry and target points (p> 0.05). The length of the trajectory was significantly correlated to the magnitude of error at the target (p < 0.001; r=0.37). This was found to be ascribable to a certain degree of flexibility of the electrode not to the registration method. Conclusions: These applicative accuracy data show that this registration method does not differ significantly from the one based on additional bony markers. It has the advantage of a homogeneous distribution of the markers around the whole volume of the skull. In addition, it does not require the placement of bony fiducials, which saves time. The mean error at the entry remains infra-millimetric and is consistent with the results of similar studies. The order of magnitude of the difference is not clinically relevant and derives mainly from methodological differences in the metrics for quantifying accuracy.

Dates et versions

hal-04220828 , version 1 (28-09-2023)

Identifiants

Citer

Iris Carron, Marc Zanello, Sami Barrit, Samuel Medina, Elodie Garnier, et al.. Applicative accuracy at entry and target points for robot-assisted depth electrodes implantation for drug-resistant epilepsy. Monocentric retrospective analysis of 1090 consecutive trajectories. Brain and Spine, 2023, 3, pp.101912. ⟨10.1016/j.bas.2023.101912⟩. ⟨hal-04220828⟩
31 Consultations
0 Téléchargements

Altmetric

Partager

More