Date of Award

Spring 4-2018

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Electrical and Computer Engineering


Byron M. Yu


All brain functions, from seeing and moving to thinking, rely on the interaction of multiple, functionally distinct brain areas. We know, however, very little about how different areas interact at the level of networks of neurons, or what mechanisms are used to control the routing of information through the brain. Only very recently has technology evolved to the point where we can simultaneously monitor multiple neurons in various brain areas. While such experiments enable a host of new and exciting questions about inter-area interaction, they also pose significant analysis and interpretation challenges. Here, we approach the problem of studying population-level interactions across brain areas using dimensionality reduction methods. In short, dimensionality reductions methods extract a small set of latent variables that summarize a given aspect of the data. Traditionally, these methods have been used to extract low-dimensional summaries of the population activity structure within a brain area. We propose to instead extract a set of latent variables that summarize the interaction between brain areas, i.e., instead of capturing the dominant features of the activity within an area, they capture the features that are relevant to its downstream targets. We used this approach to characterize both the population-level structure and the dynamics of the interactions between populations of neurons in two cortical areas, visual areas V1 and V2. We found that V1-V2 interactions occur through a communication subspace: V2 fluctuations are related to a small subset of V1 population activity patterns, distinct from the largest fluctuations shared among neurons within V1. We propose that the communication subspace may be a general, population-level mechanism by which activity can be selectively routed across brain areas. Furthermore, we found these interactions to be dynamic and flexible, changing rapidly under different stimulus contexts. This work thus provides a foundation for studying how multiple populations of neurons interact and how this interaction supports brain function.

Available for download on Friday, May 22, 2020