In the “People of I-LABS” series, we get to know individuals from our team. For this post, we introduce Eric Larson, a research scientist who specializes in developing analysis methods for MEG.
People of I-LABS
Vibrant, super-smart and caring: these are just a few of the qualities that describe the dozens of interdisciplinary researchers at I-LABS. Their innovative ideas and technological savviness help drive the Institute’s reputation as a world leader in child development and brain science.
And their kindness, professionalism and sense of humor greet all of the hundreds of families that volunteer each year for studies at I-LABS.
In the “People of I-LABS” series, we get to know the research scientists, post-doctoral fellows and other researchers who make up the elite team at I-LABS.Eric Larson is a research scientist working with Adrian KC Lee, an I-LABS faculty researcher. Larson also works closely with Samu Taulu, the director of the I-LABS MEG facilty. Larson develops methods for MEG, including software to analyze MEG data. He’s worked at I-LABS since January 2011 and is part of the Institute’s “brain team.”
Please tell us about yourself.
I grew up in Ann Arbor, MI. I play hockey a couple of nights a week, but the games are usually around 11 at night so I don’t sleep enough—especially since I have two young daughters.
What studies are you working on now?
I’m developing software that can help refine the brain signals that are recorded by the MEG scanner. One of the greatest advantages of MEG is that people don’t have to sit completely still in order for the machine to detect brain activity. This is particularly important for the awake and often squirmy babies that participate in I-LABS studies.
But head movements make the data “noisy” and it’s hard to see where in the brain activity is coming from. We have statistical methods to help clean up the data and zero in on the brain signals, and now we’re creating additional software to improve this signal-to-noise ratio when the brains are being scanned.
Dealing with head movements is one of the biggest challenges in the infant MEG field right now. It’s exciting to me to be working on a project that will make data collection more efficient for I-LABS researchers.
What’s the ultimate goal of your research?
We want to combine physics and signal processing techniques to enable other researchers at I-LABS to more readily answer questions about childhood development. The methods we develop will improve MEG research in both infants and adults, ultimately helping improve clinical diagnoses in the long term.
What are you most proud of in your career so far?
Becoming a maintainer of the SciPy Python package. This is an open-source software package for analyzing scientific data. It’s online and available for anyone to use for free. Most of the software is used by scientists and engineers. I’ve contributed code to it for the past few years, and in February I became one of the maintainers of the code base.
What is your most exciting memory or “ah ha!” moment from being in the lab?
When I came to understand at a deep level the physical basis of the existing denoising methods that we use. The critical components come from the work of Carl Gauss back in the 1800’s, but Samu first realized they could be applied to MEG analysis about 10 years ago. After talking to Samu, I put together the necessary fundamental bits of physics and mathematics to understand what was going on, and from ongoing conversations we then came up with some novel ideas for further denoising data.
What’s your favorite part of working at I-LABS
The MEG team is a great group of people to work with.
How has your personal life shaped your research interests?
Having two daughters under the age of 2, and seeing how quickly they develop, has made me want to understand what’s going on in there.
What are you most passionate about?
Coding. I originally taught myself to program during my time as an undergraduate, but I really dug in during my graduate and postdoctoral work once I realized we lacked some critical tools to answer our questions of interest. Now I get to spend most of my time working on those tools with excellent local and international collaborators. The people I code with say I’m particularly susceptible to nerd sniping.
What is the most defining moment for you as a scientist?
When I decided to become a research scientist instead of pursuing an academic faculty position. I want to work directly with data and algorithms as much as possible.