A comparison of electrophysiological microrecording versus automatic MR-based segmentation to determine subthalamic nucleus boundaries

TL;DR

This study compares two methods for determining the boundaries of the subthalamic nucleus during deep brain stimulation surgery for Parkinson's disease.

Contribution

The study provides empirical evidence of the concordance between MER and MR-based segmentation for STN boundary determination.

Findings

Both methods showed good concordance in determining STN boundaries (Fleiss’ kappa 0.721).

MER and MR-based methods agreed on optimal trajectory selection (Fleiss’ kappa 0.693).

The difference between MER and MR-based methods was less than 2.75mm for STN entry and exit points.

Abstract

Accurate placement of electrodes within the subthalamic nucleus is critical for deep brain stimulation (STN-DBS) in Parkinson’s disease (PD). Our objective was to compare microelectrode recording (MER) and an automatic MR-based segmentation tool (BrainLab ElementsTM) for STN targeting and the determination of STN boundaries.

Seventy-eight PD patients were included. Electrode placement within the STN and STN entry and exit points were determined by both methods and compared for concordance.

Of 344 trajectories, 269 were inside the STN, with good concordance of both techniques (Fleiss’ kappa 0.721, [95%CI 0.623, 0.819]). Concordance of MER and MR-based for the selection of the optimal trajectory was good (Fleiss’ kappa 0.693, [95%CI 0.578, 0.808]), with less than 2.75mm difference between MER and MR-based for the STN entry (upper limit of agreement 2.752 [95%CI 2.365 to 3.138] mm; lower limit of agreement -2.406 [95%CI -2.793 to -2.020] mm) and exit points (upper limit of agreement 2.750 [95%CI 2.351 to 3.149] mm; lower limit of agreement -2.577mm [95%CI -2.976 to -2.178]).

We demonstrated that MER and MR-based segmentation have a good concordance to determine STN boundaries during DBS surgery.

Figures