Predictive Computational Biophysical Modeling for Enhanced Image-Guided Functional Neurosurgery

Abstract

DBS is an effective surgical intervention for patients with movement disorders, e.g. Parkinson’s Disease (PD). The quality of DBS therapy is highly dependent on the accuracy of electrode placement and the consideration of neuromodulation parameters. However, brain shift, or soft tissue deformation, during surgery can cause mistargeting in treatment delivery, thus negatively impacting outcome. In this work, we present the development of a predictive multi-physics patient-specific finite element (FE) modeling framework for enhanced image-guided functional neurosurgery. This computational modeling framework is comprised of coupled biomechanical and bioelectric models. The biomechanical modeling approach, validated with high-fidelity in vivo interventional magnetic resonance (iMR) imaging data, simulates the biophysics of brain shift associated with DBS surgery and provides brain shift prediction leveraging merely sparse intraoperative data, subsequently an updated MR image may be provided for accurate and shift-accounted targeting, enhanced surgical navigation and guidance, as well as direct visualization. The bioelectric modeling approach is integrated with the biomechanical model to provide voltage potential as well as electric field estimation, subsequently volume of tissue activation (VTA) due to neuromodulation and tractography gauging potential patient functional responses. Lastly this streamlined multi-physics framework is used to examine the impact of brain shift on neural pathway recruitment. The development of this framework lays the groundwork of a platform for potential intraoperative deployment that can complement and supplement existing clinical approaches, contributes to the field of computational modeling in image-guided neurosurgery, as well as provides possible broader impact on improving overall efficacy and longitudinal success of treatment.