Deep brain stimulation is the most effective systems approach to treat brain diseases. It has a long history of success in Parkinson’s disease and is being actively trialled for epilepsy, Alzheimer's Disease and psychiatric disorders. The most crucial factor for efficacy is the selection of stimulation location and parameters, which is currently a manual process of trial and error. There is thus a pressing need for a systematic approach to design stimulation therapies informed by individual patients' brain networks and dynamics.

​

We have a unique opportunity to develop such an approach for patients with refractory epilepsy, whose seizures cannot be controlled by medications. Currently, some of these patients are treated with resective surgery, where a brain region, diagnosed to include the epileptic source, is surgically resected from the brain. While this treatment can be highly effective for some, more than half of patients continue to experience seizures. Clearly, there remains substantial need for improving epilepsy treatment. Since these patients already have electrodes implanted deep in the brain during pre-surgical recording and stimulation, this is thus an ideal platform to develop and validate therapeutics strategies based on deep brain stimulation.

​

To achieve this, we are developing a large-scale computational model to predict the evolution of epileptic seizures personalized to each patient. Intra-cranial recordings from over one hundred locations simultaneously deep inside each patient’s brain, together with MRI, are used to personalize the computational models.