Marco Bonizzato - Designing the next generation of intelligent neurostimulation technology

Marco Bonizzato (Postdoc, U. de Montréal) Supervision : Marina Martinez, Numa Dancause, Guillaume Lajoie

Neurostimulation is a form of electronic medicine, characterized by complex spatiotemporal properties. The future of neuromodulation medicine is multi-pronged interventions with high-density neural interfaces, regulating multiple neuronal networks concurrently with high precision. This development will necessarily require an algorithmic framework to handle optimal treatment delivery, a problem way too complex to be managed manually. I employ tools such as Gaussian Process (GP)-based Bayesian optimization (BO). In an array of pre-clinical experiments, we demonstrated that GP-BO can rapidly optimize multi-channel brain and spinal neurostimulation. Our results endorse learning algorithms as a structural part of neuroprosthetic design, for maximal personalized intervention effectiveness.