By: Kim Maialetti
During brain surgery, Gail Martin wears a motion-capture glove to help researchers study the areas of the brain that control movement.
On a recent Friday morning at Pennsylvania Hospital, Gail Martin rested on an operating room table, the top of her head encased in what looked like a clear plastic bubble and her left hand fitted with a Hollywood-style motion capture glove.
Every few seconds, Martin moved her hand to mimic the gesture that appeared on a screen in front of her. Rock. Paper. Scissors.
With each correct movement, the word “success” flashed on the screen, accompanied by a ding similar to the coin sound effect in a Super Mario video game.
All the while, two sets of electrodes placed in and on Martin’s brain recorded what she was thinking with each specific movement.
The data collected in operating room 8 have the potential to revolutionize care for patients with paralysis, as researchers ask: “What if we could use machine learning to translate brain waves into physical actions and give paralyzed patients the ability to move again?”
Uniquely positioned
That’s the question Iahn Cajigas, MD, PhD, an assistant professor of Neurosurgery and Bioengineering at the University of Pennsylvania, began contemplating after a childhood friend became paralyzed in a diving accident.
“I’ve been thinking about this idea of restoring function with devices for patients with paralysis for a long time,” Cajigas said. “It’s incredible that we’re now at this cusp of new technologies that are going to facilitate understanding how the brain works and potentially translating that to our patients.”
Cajigas is uniquely positioned to undertake this research, for which he has received $300,000 in early-stage investigator grant funding from the National Institutes of Health.
As a physician scientist, Cajigas uses deep brain stimulation (DBS) to treat patients with Parkinson’s disease or other forms of tremor for which medications are no longer helpful—patients like 71-year-old Martin, whose tremors have gotten so bad she can barely walk or play with her grandchildren.
With DBS, electrodes are surgically placed in targeted areas deep inside the brain and attached to a neurostimulator implanted in the patient’s chest, like a pacemaker for the brain. This eventually allows for electrical impulses to be sent to the electrodes, stimulating changes in the brain that reduce tremors and other symptoms.
It also allows Cajigas to study the areas of the brain—in patients who volunteer to participate—that control movement.
A window into movement
Iahn Cajigas, MD, PhD, and researcher Qasim Qureshi review data to identify consistent patterns in brain activity that will enable them to predict a patient’s intention to move in real time.
Together with placement of the deep brain electrodes required for DBS surgery, a second, high-resolution electrode array is temporarily placed on the surface of the patient’s brain, over the motor cortex.
Typical cortical surface arrays have three to four sensors. The one Cajigas uses for his research is an investigational advice made by Precision Neuroscience, which contains 1,024 electrodes. Its resolution is hundreds of times higher than the next most readily available clinical substitute.
Along with this device, patients wear a special glove that senses and records movement, and other motion trackers are placed along their body.
Patients are then prompted to perform a series of simple hand movements. While they make the motions, the electrodes and sensors—deep in the brain and on the surface—work in sync to capture a picture of how the brain coordinates movement.
“It’s a very unique window—over 20 minutes—into how our motor cortex and the subcortical structures are working together to coordinate the hand movements that we use every day,” Cajigas said, “with the idea that the insights we generate from these recordings will allow us to understand that intricate coordination.”
Why focus on hands? For one thing, Cajigas explained, the region of the brain that controls hand motion is very accessible from the area where the DBS surgery is performed.
“The second reason is that if you ask patients with paralysis what function is most important to them, particularly patients that are quadriplegic, hand function rates among the most important things to get back,” Cajigas said.
Leveraging a busy clinical center
Access to the brain is necessary for the type of research Cajigas is conducting, so his role in a busy DBS clinic where patients have already opted for brain surgery is helping to make the project possible. Other studies that have less access to patients are somewhat limited.
Researchers at Penn Medicine have also worked with patients with epilepsy, who are having electrodes placed in their brain to identify the condition’s sources, but they do not sample the movement areas deep in the brain because it’s not where epilepsy occurs, Cajigas said.
Additionally, research has occurred in patients who are paralyzed, but because of the paralysis, it is difficult to study actual movement.
“In our approach, we’re actually leveraging the fact that we’re a very busy clinical center for DBS,” Cajigas said. “It’d be very challenging to study this system at all, outside of this limited context. You’re not going to find healthy volunteers who are willing to allow you to put electrodes deep into the brain, like DBS requires.”
Colleague Bijan Pesaran, PhD, the Robert A. Groff, MD Professor of Neurosurgery and professor of Bioengineering and Neuroscience at the University of Pennsylvania, agreed that the volume of patients Cajigas is able to study helps accelerate the pace of research.
“Instead of looking at one subject every few years or every year or so, we can study a different patient each week or each month,” Pesaran said.
Developing new software
Data is being collected through a micro electrode recording machine operated by electrophysiologist Melanie Donley, PhD, while Iahn Cajigas, MD, Ph,D, and Medtronic representative Paul Sherlock observe.
Pesaran’s research in sensor development and in engineering the systems needed to acquire and interpret brain signals has also served as an advantage. Together, Cajigas and Pesaran developed new software called Thalamus to synchronize and integrate the data—about 80 gigabytes from each procedure—for analysis.
What they’re looking for are the different frequency bands related to movement, how they behave, and the interplay among them. This is where artificial intelligence and machine learning come into play.
“As humans, it’s very hard for us to find patterns in large amounts of information. It is literally the needle in the haystack,” Cajigas said. “We have thousands of sensors that we’re trying to make sense of.”
“AI and machine learning techniques have really allowed us to look at the data as a whole and rely on the computer to help us identify some of these patterns,” Cajigas continued.
The goal is to identify consistent patterns in brain activity that will enable Cajigas to predict a patient’s intention to move in real time. In turn, programs can be created to identify when the brain is thinking about movement and translate that activity to a prosthetic or robot that will bypass the injury and help a patient move again.
Cajigas and the research team continue to analyze data from new patients and are aiming to have enough information for publication in the next several months.
Hope for people with paralysis
Cajigas’s cutting-edge research—taking place at the country’s oldest hospital—offers hope for people with paralysis.
“I imagine a world where, if we can understand the intricacies of hand and arm function, we can take somebody who’s an amputee and restore the movement with a prosthetic,” Cajigas said. He imagines the possibility even of stimulating and re-articulating paralyzed arm and hand muscles.
“We hope to provide patients with some autonomy and independence,” Cajigas added. “We hope to improve their lives.”
Martin hopes so too.
She recognizes how research helped make DBS available to treat Parkinson’s in people like her, allowing her to try to walk again and play with her grandchildren.
“People before me did this so I could have this type of surgery,” Martin said. “I want to be able to help somebody else.”
New tech, new possibilities
As part of the study, Cajigas uses an electrode array called the Layer 7 Cortical Interface that is engineered to record neural activity at resolutions hundreds of times finer than conventional cortical surface arrays.
With 1,024 sensors on an area of 1.5 square centimeters—about the size of a shirt button—the device enables Cajigas to see precisely how the brain coordinates complex movements.
Its design allows for safe implantation and removal by neurosurgeons without causing harm to delicate brain tissue. The device, developed by Precision Neuroscience, has the potential to help researchers like Cajigas transform the lives of patients with neurological disorders.
“The technology that we’re working on now with Dr. Cajigas is oriented ultimately to provide thought-based control of external devices and computer systems to people who are paralyzed, and that’s never been possible before,” said Ben Rapoport, MD, PhD, chief science officer and co-founder of Precision. “That’s only been really dreamed of, and it’s been dreamed of for 20, 25 years.”
Rapoport noted that while the tools and techniques used to study the brain and nervous system in the 1980s and 90s were highly useful, the advances being made today are “totally transformational.”
He added, “I think that all of us will see a new standard of care in clinical neuroscience over the next couple of years.”