A Computational Model of Epilepsy and Response to Electrical Stimulation

Implantable electrical stimulation devices are now an important treatment option for the patients who do not respond to anti-epileptic medication. We propose to use reinforcement learning, a technique wherein an agent “learns by doing” in a stochastic environment, to optimize the stimulation patterns of a closed-loop stimulation device. To facilitate this research we have created a simulation environment to enable rapid testing and evaluation of these methods.