A bit of background:
I’m originally from Berlin, and I came to Johns Hopkins as a graduate student in biomedical engineering in 1994. Now I’m a professor in the department and co-director of the PhD program. And I teach a course in neuroengineering.
A quick description of what I do:
I’m an engineer, so I try to build and tinker with stuff; and sometimes break stuff. It’s awesome. And I scribble on boards. (Laughing.)
A deeper dive into my research and the questions we’re asking:
We study the development and function of the auditory cortex. A few questions my team and I are trying to answer: How is our experience of the world shaping the way we hear, and how are we processing sound in general? How does very early sensory experience change the circuitry of our brain?
So we’re pursuing two streams of research. One is the study of the adult, how does the adult auditory cortex work? How are you listening to me right now? How are you segregating my speech from all the other noises in the background, the computer, that air conditioner, and all these kinds of things. How is your attention helping you do this, and what are the underlying circuits? Then, how is this function shaped by very, very early sensory experiences? We are very much focused on the early stages of life, which in humans would also be prenatal. So, in that case, how would the fetal experience shape your perception of the world, how is fetal and early postnatal sound experience important? What circuits are influenced by this experience?
How I came to be involved in this line of research:
I completed my master’s degree in electric engineering at the Technical University in Berlin, and I was working on a project to build chips for digital television – some of the first chips for digital television. We were testing different algorithms to compress video to use less space and to be more easily processed through telephone wires at that time. So, we were throwing away a lot of information by compressing videos. Same as what you do with digital pictures or music. You compress a picture; you throw away information. So, we could measure how much information was being thrown away, but the movies looked pretty much the same. So that got me interested in trying to understand what and how the brain is actually computing. How are we extracting information from the world? I could measure with a computer that I was throwing away information, but my perception was unchanged. So that got me thinking about neuroscience.
So, then I came to Hopkins to study biomedical engineering with a focus on systems neuroscience to work at the interface of technology and the brain. At that time, biomedical engineering was where a lot of brain electrophysiology was done because engineers could build cool equipment and have quantitative methods to deal with data. I was really interested in combining engineering with advances in physiology to study the brain at a deeper level.
That is where my interest in development started. I was working on adult animals and when you start recording from neurons, you see how exquisitely they’re tuned, responding to specific sensory stimuli. To me, that begs the question, how does the system get built? It’s so complex. How in the world does it happen? If you like to build things reliably, you make a blueprint. But where is the blueprint for the brain? You have very few genes, so it’s very complicated. So, I wanted to learn about brain development and left Hopkins to go to Harvard for postdoctoral research. I worked in Carla Shatz’s Lab to study development – she’s one of the pioneers behind our understanding of the principles of early brain development.
Image Credit: Geometric comparison of topography within and across mice / Patrick Kanold
How I approach a new research question:
A question pops into my mind, many times while I’m on a run. Then I have some coffee, and I think about how to approach that question. Next, I go to the lab, and I’m excited about it, and I talk to my students and postdocs, and I get them excited (or they shoot down my ideas), and then we go do it. (And seek funding, of course.)
Our most challenging projects:
What we’re doing right now – the holographic optogenetic stimulation of small neuron populations in behaving animals– has been the most challenging. Almost 10 years passed before we really started to get good data. We are finally getting the first data where we are manipulating small populations of neurons in behaving animals and trying to bias their perception. What we are trying to do is to essentially implant auditory information into their head and see if we can change their perception of sounds.
It was so hard and took so long because multiple things had to work together. We had to get the biology right, train animals to perform tasks reliably. Our imaging and stimulation microscope technology has to also work properly. When we started, there was no commercial microscope to do this targeted stimulation. In collaboration, we built our own microscope; and we had to learn how to manipulate light well. Then the pandemic hit, and we moved to JHU. So that was very disruptive. In the end we ended up with a commercial microscope because while our custom-built microscope worked, it was so complicated to use. By the time it was ready to use, the animal was not ready anymore. So, the setup was too complicated, which was one of the problems. We also had to optimize our ability to control cells right, optimize the viruses to deliver genes to cells to make them controllable by light, etc. What was out there didn’t work very well for us. So we had to figure out how to make our own viruses to be able to simultaneously image and stimulate cells. Just learning all of these disparate things and having the right team of people together to do this took a long time.
So, for these kinds of hard questions, you need to have teams working together that come from different backgrounds, in this case, physics, engineering, and neuroscience. And maybe some UI designers (laughing). We have many computers with many different programs running simultaneously that need to talk to each other, so it is very complex. And I think this is an example of OneNeuro, the coming together of different disciplines that might not ordinarily interact to improve our understanding of the brain.
Exciting trends:
We can now simultaneously observe thousands of neurons interacting with each other, while the animal is behaving. And we can then decode the activity of these neurons to understand what the whole network is doing and then go in and alter this activity. We could not do that before. Previously, a lot of neuroscientists relied on recording from a single neuron and trying to infer from many, many single neurons what the brain does. Imaging changed this allowing us to image many neurons but was restricted to small imaging areas in the upper parts of the cortex. Now, we can go deep and also in 3D. So we are heading towards being able to actually observe entire larger-scale networks at once, which is really fun. And then, we can combine this with molecular biological methods that can identify cell types. For example, we can image many neurons and then identify every single neuron that we’ve imaged and go and see what the molecular identity of each neuron is. Combining these techniques is great. You can then imagine doing this across the lifespan of the animal, for example, to look at brain changes on many levels at once. That’s beautiful. You can see how the brain is changing on a molecular scale, on a neural connection scale, but then also on the large-scale network scale; and then connect these things by data science. I find that very exciting. Again, this requires labs and teams working together.
Thoughts on OneNeuro:
OneNeuro is a really, really, really good idea. Neuroscience is so interdisciplinary and broad. So getting to know what everyone is doing and learning each other’s language is critical. Engineers and biologists have different languages and cultures, so it is important to learn from each other to work in teams. It is also great for students to build diverse avenues for their careers. So building community is good on many levels. It would be fantastic, for example, to have student retreats for everyone who’s doing neuro; spend a few days having talks, poster sessions, and other interactions.
In my spare time:
I run (ultra) marathons, go backpacking and scuba diving, and I like to ride motorcycles. Also, now that our daughter is away at college, my spouse and I keep busy with our pets: one dog, two conures, a cockatiel, and a parakeet.
Photo Location: On the way back from Deadhorse, when passing the Arctic Circle going South. Deadhorse is located within Prudhoe Bay in North Slope Borough, Alaska, United States. July 30, 2022, 11:11 PM.