The Virtual Epileptic Patient: Individualized whole-brain models of epilepsy spread

Individual variability has clear effects upon the outcome of therapies and treatment approaches. The customization of healthcare options to the individual patient should accordingly improve treatment results. We propose a novel approach to brain interventions based on personalized brain network models derived from non-invasive structural data of individual patients. Along the example of a patient with bitemporal epilepsy, we show step by step how to develop a Virtual Epileptic Patient (VEP) brain model and integrate patient-specific information such as brain connectivity, epileptogenic zone and MRI lesions. Using high-performance computing, we systematically carry out parameter space explorations, fit and validate the brain model against the patient's empirical stereotactic EEG (SEEG) data and demonstrate how to develop novel personalized strategies towards therapy and intervention.