This proposal is about ultrasound neuromodulation (UNMOD) for the treatment of drug-resistant epilepsy. The physical mechanisms by which ultrasound interacts with neurons are many and complex. This complexity precludes the understanding-driven optimization of UNMOD systems and protocols. Indeed, published preclinical UNMOD studies for epilepsy have tested only a very limited set of UNMOD waveform parameters. The UNMOD literature further suggests that its efficacy is sensitive to the used ultrasound protocol, underlining the future need for optimized patient-specific waveforms.

In this project, we propose a closed-loop UNMOD system that adaptively updates the ultrasonic waveform parameters with the goal of maximizing epileptic seizure suppression in the hippocampus. We first optimize the controller in high-throughput simulations with a surrogate simulation model of the hippocampal environment. Subsequently, closed-loop UNMOD experiments are performed in a mouse model for epilepsy.