Mesoscale Imaging Reveals the Markovian Dynamics of the Brain

Speaker: Jordan Culp

Date: Wed, May 4, 2022

Location: Online

Conference: Emergent Research: The PIMS Postdoctoral Fellow Seminar

Subject: Mathematics

Class: Scientific

Abstract:

With mesoscale imaging, we can optogenetically record the calcium signals from the entire cortical surface of the mouse brain. However, a mathematical analysis to assess the stability or changes to the brains dynamics remains elusive due to the size and complexity of the underlying data. Here, we apply a novel Continuous-Time Markov Chain approach to assess changes to the dynamics of the mouse brain under the application of different drugs, visual stimulation, and seizure induction. In all cases we can create a kind of dynamical bar-code of the brain dynamics of the mouse by computing Markov transition probability matrices and occupancy distributions. This dynamical bar-code is unique and reproducible for each mouse, yet changes in consistent ways as a result of our experimental manipulations. Thus, we argue that a Markovian description of the mesoscale brain is sufficient for detecting dynamical changes. In this talk, I will describe the experimental background and significance of our results, along with the derivation and detailed presentation of our mathematical model. This is joint work with the McGirr, Teskey, and Nicola labs at the University of Calgary.