MSU researcher taps computation and big data to study brain function

Dec 12, 2018

Mark Reimers is an associate professor in the neuroscience program in the College of Natural Science where he integrates statistical analysis with biological theory while analyzing and interpreting the very large data sets now being generated in neuroscience, especially from the high-throughput technologies developed by the BRAIN initiative. He's an investigator at MSU's Institute for Quantitative Health Science and Engineering.


White: Describe your research for our listeners.

Reimers: Sure. All of my research is about computation and big data applied to brain function. There are four main areas. The primary focus of my research is analysis of brain activity at a very fine level, at the level of individual cells communicating with each other in the brain. I also am working on some projects involving fine-scale, fine-grained analysis of motion. That is when we're looking at people paying attention or possibly moving in a way that may indicate neurological disorders.

We also work on genes in the brain. We're very interested in what kinds of genes help define the different types of brain cells and, in particular, what goes wrong in psychiatric illness. Are there any sort of genetic signatures or characteristics of gene expression that may signify or predispose to a particular psychiatric disorder? And, finally, we do some computational modeling. That is trying to understand the dynamics of the brain and how the brain dynamics might shift under different conditions.

White: Mark, how mature is the research, and what are some practical applications we may see out of it, and when might that be?

Reimers: All of what I'm doing right now is pretty much basic research with some long-term applications in mind. So I wouldn't say you'll see any of this coming to a clinic near you anytime next year. But we're thinking that the motion analysis, for example, could be used both to help diagnose neurological disorders in a relatively inexpensive way and also to help track the progress of autistic children under therapy. That's a project going on with the Child Development Lab right now.

The work on understanding brain function is not likely to have practical applications in the next five years but will ultimately, I think, give us a lot more insight into how to design artificial intelligences and into how various thought disorders actually occur. We know, roughly speaking, that stresses or different kinds of chemical disorders can influence people's thinking, but we don't really understand how, and we hope to be able to understand that over the next few years.

Perhaps the more immediate practical application would be if we can diagnose incipient psychiatric disorders. If you can see someone starting to have a breakdown and support them and treat them early on, then you typically have a much better outcome than if they have a full-blown breakdown.

White: So Mark, what's next as you pursue these research frontiers?

Reimers: One of the biggest challenges is actually getting data. That is, we want to be able to record from many, many, many neurons. Your brain has roughly 16 billion neurons in the forebrain and another 70 billion in your cerebellum. Even a small animal's brain that we study may have 30 to 100 million neurons in the forebrain. That's an awful lot to record from, so a lot of our effort is developing new techniques.

And then we want to understand how brain dynamics change under various conditions. We're looking at, for example, how memories are formed. So we're able to monitor the subcortical regions, the hippocampus, where a lot of the immediate traces of memory are laid down simultaneously with the cortex, where a lot of the long-term traces are stored. And we're watching the communication between those areas in an attempt to understand how memories are formed. That's a next step.

Another next step is to try to characterize what aspects of human motion indicate attention, if you're looking at autistic children, or what aspects indicate particular kinds of neurological disorders. We'd like to understand, also, what signatures we might see genomically of different kinds of psychiatric problems.

White: Mark, can you talk a bit about this collaborative ethos that Chris Contag is developing at the IQ and how your research benefits from this intersection of medicine and engineering?

Reimers: Sure. We're interested in fundamentally physiological and medical problems, but we need a lot of high tech to do this, and so this fruitful interaction between the people who have an engineering background who can help build some of the equipment we need, together with physiologists and computational people like myself who can sort of figure out how to get the maximum benefit from the equipment. I think this will lead to some substantial steps forward in the next few years.

White: Summarize then for us, Mark, what you'd like us to know about your work.

Reimers: Well, ultimately we want to understand how the brain works, and perhaps shed some light on this long-standing mind/body problem. How is it that this gray jelly in your head can give rise to thoughts, feelings, aspirations, hopes, and fears? There are many ways that all the connections among this jelly can go wrong. How can we help characterize them, and can we help treat them?

White: Well, Mark, thanks for telling us about it, and best wishes moving forward.

Reimers: Okay. Thank you, Russ.

White: That's Mark Reimers. He's a professor of neuroscience in Michigan State University's College of Natural Science, and he's part of the IQ team at MSU. That's the Institute for Quantitative Health Science and Engineering

MSU Today airs Sunday afternoons at 4:00 on 105.1 FM and AM 870.