Faculty Tech Talk: Treating the Living Brain

What’s it like to decode the inner workings of the human brain? It’s something akin to “trying to fix a car when that car is so incredibly complex you can’t even figure out how to open the door,” says Professor Elizabeth Hillman.

The immense complexity of our grey matter has long inspired and mystified scientists. But one day soon, new imaging tools pioneered by Hillman and fellow Columbia Engineering Professor Elisa Konofagou could transform what’s possible. For instance, Hillman recently developed microscopy technology that offers unprecedented real-time 3D imaging of living organisms—like the firing of individual neurons—allowing scientists to conduct entirely new kinds of research. At the same time, Konofagou’s lab is pioneering new techniques for precisely targeted ultrasound therapeutics that could trigger the immune system to attack brain tumors or reverse neurogenerative diseases by moving therapeutics through the blood/brain barrier.

The two professors recently sat down with Dean Mary C. Boyce and a gathering of students at Carleton Commons to discuss their cutting-edge efforts to understand and heal brains. It was the latest in a new series of talks connecting students with campus researchers confronting extraordinary challenges.

Hillman, a professor of biomedical engineering and radiology, is director of the Laboratory for Functional Optical Imaging, as well as a principal investigator at Columbia’s Mortimer B. Zuckerman Mind Brain Behavior Institute. Konofagou, also a professor of biomedical engineering and radiology, heads the Ultrasound Elasticity Imaging Laboratory.

In a lively and wide-ranging discussion, the professors touched on topics ranging from the physics of brain imaging to the role of artificial intelligence. We’ve excerpted a few highlights, condensed and edited for clarity below:

Q: You each engage in very cross-disciplinary collaborations with neuroscientists and clinicians. How have those interactions influenced your work?

Elisa Konofagou: One reason why I work with the brain is because neurologists came to me and said “we need to have more kinds of tools available to us.” There’s a lot of desperation among clinicians because Alzheimer’s, Parkinson’s, ALS, brain tumors, you name it, are galloping ahead… Clinicians have the urgency of treating the brain and diagnosing it properly, and all these diseases are very undertreated if not untreatable right now. Neuroscientists are [more] interested in understanding the normal brain and how it works, because there’s still a lot we don’t understand about it. But this is actually the bottleneck for a lot of us—how do you actually go inside the human brain and probe it and do cool stuff without damaging it? We don’t have the tools we want right now.

Elizabeth Hillman: I would say one thing that interdisciplinary collaborations have forced me to do is go outside my comfort zone into new areas, where you see something and you say, “I have no idea what this is… but I am going to figure it out” Actually, I want to blow the horn of biomedical engineering here, because in BME we are these integrators of information. We have to do this to keep moving forward and discovering. If you just say, “I do imaging” and you just do imaging, that’s no good. When you have the privilege of seeing these things for the first time, it’s your responsibility to follow up. The challenge of understanding and treating the human brain is absolutely one of the biggest things going on right now, and it really takes input from all different fields.

One exciting new cross-over is neuroscience and data science. In our case, we can generate 1.5 terabytes an hour of imaging data, and we have to turn that into knowledge. We need an influx of people who understand machine learning, and deep learning models to garner meaning from the kinds of data and images that we can now get – and to ultimately to decode how the brain drives behavior.

Q: In terms of ultrasound, we think of it as an imaging technology, but it’s also now shown to be therapeutic. Can you talk a little about that?

Konofagou: There’s a Nature paper from 1905 on focusing ultrasound waves to knock down bees. We know how ultrasound can have bioeffects; mechanically and thermally, it can change things. We’ve come a long way…and now therapeutic ultrasound is in the clinics. If someone has tremors as part of their Parkinson’s disease, normally you have to use DBS [deep brain stimulation], which is basically a pacemaker in the brain. It’s major surgery. You have to insert electrodes all the way to where the midbrain is and have to have a permanent device that’s based in your brain. But with ultrasound you just go in for 10 minutes. It’s guided by MRI, and they ablate the brain in very small regions and the tremors stop.

We’re also understanding that between burning the brain and imaging the brain, there’s a huge other range within which you can navigate to cause minor mechanical effects—like we cause with stimulation—or minor thermal effects—where you can facilitate chemotherapeutic drugs to go [into] a tumor. You can also trigger an immune response. If you focus your ultrasound beam on the tumor, the T cells will migrate to the tumor and try to stop it from metastasizing. There are a lot of new techniques in therapeutic ultrasound, we just have to make sure it works in humans.

Q: We’ve seen tremendous progress over the last five years around research and understanding of the brain, yet we also know that we haven’t scratched the surface. Where do you think we’ll be in ten years?

Hillman: Things have really accelerated. The number of research tools that are now available are incredible and at the fingertips of so many more people, including people who do disease research, which is a great transition to see… although truly ‘understanding the brain’ is going to take a long time. However, I am excited about another angle too, in which we are starting to appreciate how and why things that you wouldn’t normally call normal medicine are effective for things like depression or even brain diseases. For example, why do you feel better if you go and do exercise?—we are starting to understand that these effects are biological interactions that are meaningful. I think progress in the near term will be made in lot of ways, for example, I think we will find that therapy in an autistic child is actually remodeling the brain and causing it to be improved at a cellular level. We’re already learning all of this now… but I think it’s going to be incremental – I don’t think we will ever see some big announcement that “now we understand the brain,” but these little pieces will come together and start to make sense and guide us toward better diagnostics and treatments for human brain health.

— Jesse Adams, Columbia Engineering

About garen