All you need is oscillations
A simple oscillatory model goes a long way…
Updated: 2022-01-17
I’ve been in Marseille now for a few months and it’s been an amazing experience. Part of what I’m working on here is models of the brain but I’m also working on the inverse: dynamical models and the role that control theory can play in understanding and changing the behavior of those dynamics models.
This last bit is important for the future of DBS but the field isn’t quite there yet. Recent work from the lab of Dr. Dani Bassett at Penn is spearheading the introduction of control theory into the field of neuroimaging and neuroscience. It’s very exciting to see and opens up an amazing world of research since both neuroscience and control theory are deep, rich fields.
I got an introduction to control theory during my Master’s in ECE at GT. I took Network Control and Nonlinear Control with Dr. Magnus Egerstedt and am in the middle of watching some of his Optimal Control lectures. It’s an absolutely amazing time to get into control theory because it’s actually the natural next frontier of engineering now that machine learning has come in and “spanned” the field of statistical signal processing. To be clear, there’s still a lot of work to be done in machine learning, but machine learning is fundamentally concerned with learning relationships between variables. Controlling those variables is a whole different beast that requires us to learn a whole different set of rules.
And learn we shall. Hop on over to my main repository for my control theory project during my Whitaker Fellowship: [https://github.com/virati/KuramotoDBS]. Special thanks to Rohit Konda for a lot of the fundamental work done to get this project to a stage where it’ll be (more) straightforward to finish in the mini-est of mini-PostDocs.
-V