Elsevier

Brain Stimulation

Volume 10, Issue 1, January–February 2017, Pages 46-50
Brain Stimulation

Role of Soft-Tissue Heterogeneity in Computational Models of Deep Brain Stimulation

https://doi.org/10.1016/j.brs.2016.09.001Get rights and content

Abstract

Background

Bioelectric field models of deep brain stimulation (DBS) are commonly utilized in research and industrial applications. However, the wide range of different representations used for the human head in these models may be responsible for substantial variance in the stimulation predictions.

Objective

Determine the relative error of ignoring cerebral vasculature and soft-tissue heterogeneity outside of the brain in computational models of DBS.

Methods

We used a detailed atlas of the human head, coupled to magnetic resonance imaging data, to construct a range of subthalamic DBS volume conductor models. We incrementally simplified the most detailed base model and quantified changes in the stimulation thresholds for direct activation of corticofugal axons.

Results

Ignoring cerebral vasculature altered predictions of stimulation thresholds by <10%, whereas ignoring soft-tissue heterogeneity outside of the brain altered predictions between −44 % and 174%.

Conclusions

Heterogeneity in the soft tissues of the head, if unaccounted for, introduces a degree of uncertainty in predicting electrical stimulation of neural elements that is not negligible and thereby warrants consideration in future modeling studies.

Introduction

Magnetic resonance image (MRI)-based bioelectric field models of the human head show promise for optimizing deep brain stimulation (DBS) therapies [1]. These models typically use finite element methods to solve for the electric field generated during DBS and are coupled to cable models of axons to quantify the neural response to stimulation. However, a wide range of different model representations for the human head are currently used in both academic and industrial research. We hypothesized that these volume conductor differences are responsible for substantial variance in model predictions.

Recent studies have identified the degree of heterogeneity and anisotropy that is required to accurately model the electric field in the human head [2], while others have highlighted the importance of accurately representing boundary conditions in models of DBS [3], [4]. Therefore, the goal of this study was to integrate the latest advancements in anatomical and electrical DBS modeling to identify the role of cerebral vasculature and soft-tissue heterogeneity outside the brain on the neural response to stimulation. We used a highly detailed multimodal image-based anatomical model of a human head and neck, or MIDA, which was recently made available by the FDA [5], as a template for constructing a range of subthalamic DBS models. The degree of model detail ranged from complex to simple and enabled us to quantify errors in predicting the neural response to DBS that are associated with ignoring cerebral vasculature and soft-tissue heterogeneity outside the brain.

Section snippets

Materials and methods

The methodology in this study was adopted from our previous work [2] with specific integration of the MIDA atlas volumes [5].

Results

Electrical loads predicted in the base case, MIDA12, were 1.26 kΩ and 1.61 kΩ in the monopolar and bipolar stimulation cases, respectively, which are representative of clinical measurements [20], [21]. Stimulation thresholds for evoking a muscle twitch in subthalamic DBS are typically around 5 V [22]. MIDA12 predicted that between ~0% and 45% of DCF axons were activated with monopolar stimulation at 5 V and 10 V, respectively, also consistent with a previous study [23].

Altering heterogeneity

Discussion and conclusions

The goal of this study was to determine if the cerebral vasculature and heterogeneity in the soft tissues outside of the brain could be ignored in computational models of DBS. Replacement of blood vessels with other tissues had a negligible effect on model predictions. However, combination of all the soft tissues into a single homogeneous and isotropic medium, as is commonly done in DBS models, had a non-negligible effect that should not be ignored, especially if attempting to generate

Acknowledgments

This work was supported by an Individual Postdoctoral Fellowship Grant (F32 NS096839) and Research Grants (R01 MH102238; R01 NS085188) from the National Institutes of Health, and made use of the High Performance Computing Resource in the Core Facility for Advanced Research Computing at Case Western Reserve University. We would like to thank Kabilar Gunalan for his assistance with the models.

References (27)

  • M.C. Keuken et al.

    Ultra-high 7T MRI of structural age-related changes of the subthalamic nucleus

    J Neurosci

    (2004)
  • E.O. Richter et al.

    Determining the position and size of the subthalamic nucleus based on magnetic resonance imaging results in patients with advanced Parkinson disease

    J Neurosurg

    (2004)
  • CIBC

    Seg3D

    (2015)
  • Cited by (0)

    View full text