Abstracts

Rationale: Resective surgery is one of the few treatment options available for patients with pharmacoresistant epilepsy. The goal of these surgeries is to eliminate seizures with minimal loss of brain function. Prior to surgery, subdural electrode grids are implanted to map epileptic and functional areas. This method has been successfully used for many years, but little is known about the interactions between the applied electric field and the response of cortical neurons. However, important clinical decisions rely on implied correlations between the electrode location, the estimated spread of stimulation, and the cortical tissue planned for resection. In turn, there is a need to improve understanding of the neural response to cortical stimulation.Methods: We addressed this question with an interactive system that was developed to predict the volume of tissue activated (VTA) based on patient-specific models of the electrode, electric field and neural response to cortical stimulation. The system uses the finite element method (FEM) to create a discretized brain volume with cortical surfaces constructed from MRI, electrode positions determined from CT, and electrical conductivity tensors determined from diffusion tensor MRI (DTI). VTAs were determined from simulations where the electric field from the stimulating electrodes was imposed on populations of pyramidal cell model neurons placed within layers 5/6 of cortex. VTA estimates were defined based on the position and orientation of cortical electrode(s) and the stimulation waveform within a realistic, patient-specific brain volume.Results: In a retrospective study of a 20 year old patient we found that the position of the electrode and amplitude of stimulation both had important modulatory effects on the VTA, which was dependent on the proximity of the electrodes to the cortical surface and the thickness of the cerebrospinal fluid layer beneath the electrodes. Both of these factors were correlated with the impedance of the electrode, which may be useful for parameterizing models in multi-patient studies. Further, the inclusion of realistic cell models and DTI-based conductivities each had a strong effect on the VTA as indicated by its complex, non-spherical shape. Model validation is currently underway in a prospective study with clinically measured responses during stimulation of sensory and motor cortex.Conclusions: Our long-term goal is to use these methods to improve the accuracy and efficacy of cortical mapping with subdural grid electrodes, and to enhance understanding of the interaction between cortical stimulation and physiological responses for epilepsy patients.