A method for electric field simulations and acceleration measurements for intraoperative test stimulation
Abstract
Introduction:
Despite an increasing use of deep brain stimulation (DBS), the fundamental mechanisms underlying therapeutic and adverse effects and the optimal stimulation site remain largely unknown. Computer simulations of electric entities such as electric field or current density are increasingly used to try to identify the stimulated volume around implanted DBS electrodes. So far no group has considered simulations for intraoperatively obtained test stimulation data. The aim of the present paper is to propose a methodology allowing an optimal exploitation of test stimulation data and of the clinical outcome quantitatively assessed by acceleration measurements. By performing patient-specific electric field simulations for stimulation amplitudes at different anatomical positions we aim at getting supplementary data about implicated structures and the mechanism of action of DBS. In order to illustrate technical and clinical feasibility, the presented methodology has been applied to one patient.
Methods:
One patient with essential tremor presenting tremor and bilaterally implanted in the ventro-intermediate nucleus (Vim) has been included in the present study. Vim and its anatomic neighbors were preoperatively manually outlined using the iPlan software (Brainlab, Feldkirchen, Germany) according to spontaneous MRI contrasts. The identified structures were exported via a specifically designed interface based on VVLink (Brainlab, Feldkirchen, Germany) and VTK (VTK 5.2.0, Kitware Inc. Clifton Park USA). During the intervention, intraoperative test stimulations were performed in 8 to 9 positions per trajectory and on four trajectories (two per hemisphere). Tremor was recorded using a 3-axis accelerometer at each stimulation position just before the start of test stimulation (=baseline) and during the tests. Changes in tremor were expressed as percentage improvement compared to baseline. For each stimulation position, two stimulation amplitudes were chosen for electric field simulations, one with no or low clinical improvement and one with high improvement. A finite element method was applied to calculate the electric field distribution. Conductivity values were deducted from the patient's T1 weighted MRI. An isofieldlevel of 0.2V/mm was chosen and the points of the isosurface were exported. Isosurface, extracted anatomical structures and trajectories were visualized together. The percentage and the numberof appearance of each structure inside the isosurface were calculated and respectively noted. The number of appearance identified with the simulation based approach and from the classical approach where only the anatomical position of the center of the measurement electrode is considered, were compared. A correlation analysis was performed between the clinical change and the percentage of the structure covered by the electric field, taking into account the data of all positions.
Results:
69 electric field simulations were performed in total for the four trajectories. Structures identified at least once inside the isosurface of the electric field were the intermediolateral (InL), the dorsolateral (DL), the ventrooral (VO), the Vim, the ventrocaudal medial (VCM) and the center median (CM) nucleus. When comparing the numbers of appearance of each structure between the simulation based and the classical approach, the VO and VCM appeared more often and the DL only appeared with the simulation based approach. The highest improvement was obtained when VO, VCM and CM were present inside the stimulated volume. The correlation of the percentage improvement with the percentage of structure included in the isosurface showed that in some structures (VCM, CM, DL, VO) the clinical improvement varied a lot without a significant change in the percentage of structure volume included. In the InL and Vim an increase of structure parts might correlate with an increase in clinical improvement.
Conclusion:
A workflow and methodology making possible electric filed simulations on manually outlined anatomical structures could be established. This new concept will allow the analysis of a high amount of intraoperative data obtained in a clinicalstudy which might help to elucidate the mechanism of action of DBS. First results seem to confirm published data hypothesizing that the stimulation of other structures than the Vim might be responsible as well for good clinical effects. But the analysis of more data is necessary to draw any final conclusion.
Origin | Files produced by the author(s) |
---|
Loading...