The Feasibility of Methods for Solving the Inverse Problem in Electrocortigraphy Data Analysis

ECoG is a neuroimaging modality recording local field potentials produced by neural populations using electrodes placed directly onto the cortical surface. Supeficially located ECoG recording sites can’t capture the activity of sources located in cortical gyri. To partially recover the activity of individual deep sources and better interpret the ECoG data, one must solve the inverse problem. Here, we have developed a pipeline to fuse multimodal data comprising of MRI , CT, and ECoG data for forward modeling and explored the feasibility of EcoG-based source reconstruction with local and global inverse solvers, evaluating their performance on the basis of simulated and real data obtained from motor-paradigms. We have also compared finger trajectory decoding accuracy from ECoG data and based on the inverse solution. In addition to the goal of developing advanced neural prostheses for paralyzed people, the work on ECoG-based cortical mapping contributes to clinical diagnostic methods.