(18) Design Considerations for Epidural Cortical Stimulation: Insights from Computational Modeling

Introduction: EXternally Powered Epidural Cortical Stimulation (XCS) using a millimeter-scale, digitally programmable over-brain therapeutic (DOT) device has recently emerged as a promising treatment for neurological disorders. The device must provide focal electric fields in desired cortical regions while reducing potential side effects by minimizing current spread to the scalp. However, the impact of device design choices on electric field generation remains poorly understood. Computational modeling offers a valuable approach for optimizing device design by enabling exploration of wide parameter spaces. We employed finite element modeling to calculate electric field distributions when varying anatomical properties and device architectures.

Methods: We simulated electric field distributions generated by the DOT in a spherical head model using COMSOL Multiphysics. We placed the DOT in a burr hole and varied the burr hole properties and device design, quantifying current density and electric fields in gray matter and scalp to assess the feasibility of achieving effective brain stimulation while avoiding potential undesirable scalp side effects.

Results: Our goal was to optimize device design and devise strategies to enhance gray matter current density while minimizing scalp currents. Device placement in the burr hole and electrode geometry had a strong influence on focality and depth of stimulation, with larger electrode diameters reducing energy consumption, but maintaining scalp current levels and reducing gray matter current density. A burr hole cap was essential to prevent current leakage to the scalp, and variations in skull thickness necessitate tailored caps and spacers with optimal positioning to minimize scalp currents. Burr hole properties, including size and the conductivity of material surrounding the device in the burr hole, played a critical role in determining current flow. When filled with highly conductive fluid (edema or cerebrospinal fluid), a larger burr hole substantially reduced scalp currents and energy consumption, but led to a modest decrease in cortical current density. Conversely, using lower conductivity materials in the burr hole increased cortical current density but at the cost of higher scalp currents and energy use.

Conclusion: Device design features such as electrode geometry and cap design, in addition to burr hole properties, can be selected to maximize stimulation of cortical brain regions, minimize scalp currents, and control energy use. By quantifying the influence of various device design features and burr hole properties using computational models, we can guide rational design of XCS devices capable of successfully modulating cortical targets for therapeutic effects.