It was a historic milestone, reported around the world, when it happened about a decade ago: Using the new investigational brain-computer interface known as BrainGate, two people with tetraplegia – stroke victims suffering paralysis in all four limbs and the torso – were able to move a prosthetic arm and perform basic tasks using only their minds.
The BrainGate demonstration had been years in the making. By 2002, a team of Brown University neuroscientists had developed the system’s basics: Using a tiny electrode array the size of a baby aspirin, implanted in the brain just beneath the skull, they detected the electrical signature of neurons firing in a specific area of the brain and translated these impulses into signals that activated and controlled an external device. Together with a spinoff company from the Brown lab (Cyberkinetics), investigators from Brown, Massachusetts General Hospital (MGH), and the Providence VA Medical Center (PVAMC) began clinical research with the new technology in 2004, and showed that a person with tetraplegia could control a computer cursor on a screen and perform simple tasks such as opening emails, operating a television, and opening and closing a prosthetic hand.
The BrainGate trials combined the expertise of dozens of people throughout the VA and academic research communities, and other VA clinicians and researchers were using some of the same core technologies – electroencephalography (EEG), magnetic resonance imaging (MRI), and computational neuroscience, for example – to inform and validate their own investigations. It seemed natural that this knowledge and capability should be gathered, under the umbrella of VA’s Rehabilitation Research and Development (RR&D), into a single center that could leverage its impact.
In June 2012, with RR&D Service funding, the Center for Neurorestoration and Neurotechnology (CfNN) was formed as a collaboration between the Providence VA Medical Center, Brown University, Butler Hospital, Lifespan (Rhode Island Hospital’s health care system), and MGH. The center’s director is Leigh Hochberg, MD, PhD, one of the original BrainGate researchers, a Brown University professor, Harvard lecturer, and clinician who directs MGH’s Center for Neurotechnology and Neurorecovery. “The mission for CfNN,” Hochberg said, “is to discover, design, develop, and deploy novel neurotechnologies and other device-based therapies that will advance the rehabilitation of veterans and others with impairments in mobility, communication, mental health, and limb function.”
The work of the CfNN’s clinicians and researchers is supported by a core of technological and administrative expertise: Experts in the Neuromodulation and Imaging core make advanced software and hardware available to investigators who need the highest-quality MRI. Within the Recording, Decoding and Computational Neuroscience core, specialists record and analyze brain signals to reveal how the brain works, and how device-based therapies might be refined to modulate neural networks. A third core – Assessment, Outcome Measurement, and Implementation – is dedicated to mastering the protocols and regulatory environment for investigators and project coordinators, while facilitating access and continuity of care for study participants.
“We provide these core competencies and services not only to CfNN researchers,” Hochberg said, “but also make them available to researchers throughout the Providence VA and the national VA community.”
At the CfNN, this expertise supports the work of investigators in three focus areas: Communications and Mobility; Affective and Cognitive Health; and Limb Function.
COMMUNICATION AND MOBILITY
Investigators in this focus area aim to restore function and independence to veterans with impairments due to ALS, spinal cord injury, stroke, seizure disorders, or other disorders. With the help of new neurotechnologies and medical approaches, CfNN researchers are working to discover and apply new ways to harness the neural activity that can be recorded in these conditions. Led by John Simeral, PhD, a bioengineer at the center and an assistant professor (research) at Brown University, these activities include continued study and development of the BrainGate system.
The ongoing BrainGate trial is enabling research participants with tetraplegia to browse the internet, chat online, compose and send emails, and control televisions and other home appliances or assistive devices; in some studies, people continue to explore possibilities in using the neural interface to command robotic and prosthetic limbs. In collaboration with experts at Case Western University and the Center for Functional Electrical Stimulation (FES) at the Cleveland Stokes VA Medical Center, the team reported in the journal Lancet that they could link the BrainGate interface to an implanted FES system that delivers what are essentially artificial motor nerve impulses to muscles in the arm and hand of a person with tetraplegia. This is an exciting discovery: It suggests that people, years after paralysis, may be able to use technology to reach and grasp for objects using their own arms.
The ultimate goal for investigations of neural interfaces, said Hochberg, is to give veterans a solution they can use by themselves. In its current form, the BrainGate system conveys neural signals using cables connected to amplifiers and decoding computers that require an expert attendant. “There are all kinds of advantages to testing devices in a well-controlled laboratory,” he said, “but it’s not where these devices need to work. They need to work at home – or wherever that person wants to be in a given moment.” As a result, since the research began, that’s where the BrainGate team has been developing and testing their system – in each research participant’s home.
CfNN investigators are working to improve the reliability and portability of the BrainGate system – including the development of a wireless system first developed at Brown University, with the first human use of the external wireless system just published by a team led by Simeral. “We’ve been able to replace that big cable with a wireless transmitter … and a wireless lead that beams all of these neural data,” Hochberg said. “That’s an important step toward what will be the next generations of this research with fully implanted systems.”
CfNN researcher and PVAMC staff neurologist David Lin, MD, also a neurologist at MGH and instructor at Harvard Medical School, is exploring additional possibilities for brain-computer interface technologies, beyond assistance in everyday function for people with paralysis – they may also have the potential to rehabilitate native function by leveraging neuroplasticity, or the brain’s ability to change and repair itself.
The first step in understanding how this might work, Lin said, is learning more about how the brain changes after an acute neurologic injury such as stroke. “We are tracking the natural history of motor recovery in people with arm weakness after stroke,” he said. “From the acute stroke period forward, we perform a series of outcome measures paired with neurotechnological assessments such as MRI, EEG, and TMS [transcranial magnetic stimulation] to examine how structures and pathways in the brain that have been affected by stroke change and allow for arm motor recovery.”
Leveraging insights from these investigations, Lin is also leading a collaborative study involving MGH, Harvard, and the Providence VA to engineer new technologies to help people recover arm function lost after a stroke. “Our aim is to leverage neurotechnologies that can be deployed in the clinic, such as EEG-based brain-computer interfaces, to restore arm function,” he said. A patient who has suffered a stroke and can’t move their arm, Lin explained, can still imagine moving it – “And when they do that, there are characteristic EEG signals that are evoked over the brain, primarily in areas of the brain that control movement, like the motor cortex.” These patterns of brain activity can be linked to a robotic device that can move the person’s arm for them, restoring the connection between intended and actual movement.
“In contrast to a system like BrainGate, where the brain-computer interface can be used to bypass the injury and reenable functional activities,” said Lin, “here we’re trying to restore neural connections themselves. The idea is that after using the EEG brain-computer interface for a number of sessions, the participant who has had the stroke will actually regain lost function.” Lin’s ultimate goal is to translate these technologies to maximize functional recovery for veterans and others with stroke.
EEG plays a key role in the work of W. Curt LaFrance, Jr., MD, MPH, a staff physician at the Providence VA and a professor of psychiatry and neurology at Brown. As the clinical lead for the Tele-Seizures Clinic at the VA National TeleMental Health Center, LaFrance treats patients with neuropsychiatric disorders including seizure, traumatic brain injury, and disorders of cognition, mood, anxiety, and movement.
LaFrance’s work for the CfNN builds on the expertise of seizure monitoring units at the VA’s 17 Epilepsy Centers of Excellence (ECOE), which can characterize seizures using video synched with an EEG of brain activity. Video EEG is a key tool in making an important distinction, LaFrance said: “In many cases, if there’s abnormal brain cell firing on the EEG with a seizure, we can say: ‘That’s an epileptic seizure.’ If there are no abnormal signals, what we call epileptiform discharges, during the event, then we say, for a number of presentations, ‘Ah, that’s consistent with a non-epileptic seizure.’ And it’s so important to be able to distinguish, because anti-seizure medications don’t treat non-epileptic seizures.”
LaFrance has helped to develop and validate a course of psychotherapy to treat patients with these seizures, including psychogenic non-epileptic seizures (PNES), in 12 weekly onehour sessions that can be delivered in person or online. While he considers video EEG to be the gold standard in making the distinction between epilepsy and PNES, he has the same easeof-use concerns as the CfNN’s other investigators. “Not every veteran can spend a week in the seizure monitoring unit,” he said. “And not every veteran who is in the seizure monitoring unit actually has one of their typical events. So the study may be inconclusive.”
LaFrance’s group is experimenting in the use of a wearable wristband that can record biodata such as skin temperature, heart rate, and electrodermal activity. The hope, he said, is that these data may collectively be used to capture seizures, and to help distinguish between epileptic and non-epileptic seizures. Using algorithms and machine learning to analyze signals from the wristband, the group has identified a possible signature that differentiates between epilepsy and PNES – but it needs to be replicated among a larger sample. “Identifying novel diagnostic tools is one of our ongoing areas of focus,” LaFrance said.
AFFECTIVE AND COGNITIVE HEALTH
The work of investigators in the CfNN’s second focus area, Affective and Cognitive Health, involves devices and technologies that alter the brain’s complex neural circuitry. Noah Philip, MD, the psychiatrist at the Providence VA who directs this area, views its mission simply: “What it boils down to,” he said, “is trying to figure out how we can help people, knowing what we know about the brain, without using pills and needles.”
CfNN researchers in this focus area use imaging technologies to identify therapeutic target areas in the brains of people with neurobehavioral disorders – depression, PTSD, suicidal thinking, chronic pain, or obsessive-compulsive disorder (OCD) – and apply electromagnetic fields to increase or decrease neuronal activity in the brain.
A neuromodulation technology often used by these investigators is transcranial magnetic stimulation (TMS): an electromagnetic coil, placed against the patient’s scalp, painlessly applies a magnetic pulse that stimulates nerve cells in the brain. TMS has been shown to be effective in treating depression – though the biology of why it works, exactly, isn’t completely understood. Ongoing investigations at CfNN have suggested that specific patterns of connectivity in the brain’s neural circuitry can predict a patient’s response to TMS. The center’s experts in computational neuroscience have enabled and supported studies demonstrating that the patients most likely to respond to TMS treatment can be identified with the help of machine learning.
A team of researchers from the center and Brown University including Philip; Jennifer Barredo, PhD; Yosef Berlow, MD, PhD; Melanie Bozzay, PhD; Hannah Swearingen, BA; and CfNN Associate Director Benjamin Greenberg, MD, PhD, recently published a paper demonstrating the promise of TMS in treating suicidal ideation – but the underlying neural mechanisms of suicide remain poorly understood. Researcher Jennifer Barredo, PhD, has worked to better understand the neurocircuitry of suicide and use neuroimaging to identify those at risk for suicide. HSRD investigators Jennifer Primack, PhD, and Noah Philip, MD, co-led a recently launched first study, in conjunction with CfNN investigators to combine brain stimulation and psychotherapy to reduce suicide in high-risk veterans.
Another method of neuromodulation being evaluated at the CfNN is transcranial direct-current stimulation (tDCS), the application of positive or negative electrical impulses by electrodes attached to the scalp. Investigators led by Philip and Mascha van’t-Wout Frank, PhD, have demonstrated that tDCS, coupled with a variation of exposure therapy delivered via a virtual reality headset, can alleviate PTSD symptoms.
“One of the ways we understand post-traumatic stress now,” Philip said, “is that it’s really an imbalance in different neural networks: regions of the brain that act together to have a certain function.” In PTSD, explained Philip, a cluster of brain regions known as the salience network – because it helps the brain figure out what’s most relevant in a given moment – becomes very active when a person is under threat.
If this fight-or-flight response is activated frequently, or with unusual intensity, Philip says, “that system can get stuck in the on position – or the gas pedal, if you will, can get stuck down. What we’re doing is helping to get the brake working again.” Finding that braking system – in an area of the brain known as the dorsolateral prefrontal cortex, which is involved in self-control and regulating thoughts and emotions – was an effort supported by CfNN’s core experts and their EEG and functional MRI scans of active brains.
A recent study of veterans with PTSD targeted this area of the brain using a TMS paradigm known as intermittent theta-burst stimulation (iTBS): short bursts of magnetic waves oscillating at fives cycles per second. “That frequency, we have reason to think, helps a lot with getting the brake to work,” Philip said – it matches the frequency of neural impulses in the hippocampus, where memories of trauma are processed.
“What we have seen,” said Philip, “is that if we do this gently enough, and with enough energy, people get better.” A group of about 50 veterans participated in the study, which was a doubleblind trial comparing the results of those who received iTBS in 10 sessions over 10 days, and those who received a placebo or “sham” treatment. On a range of outcomes, those who received iTBS did better than those in the control group, both two weeks and one month after the treatments.
Further investigations of this promising new treatment will focus on optimal dose of stimulation. For veteran patients, iTBS has a major advantage over traditional transcranial magnetic stimulation: It can be delivered in three – to 10-minute sessions, compared to the traditional 45-minute TMS sessions. “That means we can provide a lot more access to care,” Philip said.
Philip’s group has now compiled data on about 800 patients who have received TMS treatments across the VA – and the data affirm that it works. “We have reason to believe it’s safe and efficacious both for depression and for post-traumatic stress.”
The study of tDCS in conjunction with virtual reality exposure is a sign of things to come for investigators in Philip’s group: joining proven clinical interventions, such as exposure therapy, with new neuromodulating technologies, some of which may be safe for people to use at home. “My primary goal is helping people get better,” said Philip. “My other goal is to understand how the brain works, so we can develop new ways to help people get better.”
Investigators in the CfNN’s third focus area seek to maximize the benefit of existing technologies – with a focus on prostheses – for veterans who have lost all or part of their upper limbs.
Linda Resnik, PT, PhD, a research career scientist at the Providence VA and a professor at Brown, directs this focus area using insights acquired over a nearly 20-year career in physical therapy. When the advanced robotic prosthetic known as the DEKA arm became available for study in the early 2000s, she was principal investigator for a multi-site VA study aimed at optimizing its design and usefulness. Her team subjected the prosthesis to an exhaustive evaluation, reporting on its features and functionality in all available configurations, and the study was used to support FDA approval of the device. Her team compared two methods of controlling the arm – now known as the LUKE arm – using either a wireless foot control system or pattern recognition software that decoded the electric signals of arm muscles. Resnik also led a study investigating the feasibility and benefits of using the prosthesis at home.
At the time, Resnik began to realize she didn’t know as much about the prostheses people were using outside a VA laboratory. “There hasn’t been enough research on currently available devices and their features compared to each other,” she said. “We don’t understand who might best benefit from which type of device. So our motivation is to get more data, so that when we have these more advanced devices, we have good comparison data to understand the ways they may differ.”
Finding answers to that question may seem simple, but Resnik’s group is careful, when designing investigations, to remember the most important determinant. Differences among people – their lifestyle needs and desires and thresholds for comfort or pain – may matter more than hardware.
The Limb Function group at the CfNN is currently involved in two studies of veterans and active-duty service members recruited from among upper-limb amputees nationwide. The first involves nationwide survey and in-person data collected with the help of several partners: four VA sites, the Department of Defense’s Center for the Intrepid at Brooke Army Medical Center, and the University of Massachusetts Medical School. Recruitment for this study is complete, the data has been collected, and multiple papers have been published. The team is still analyzing those important data. Some issues revealed so far, Resnik said, have implications for further research in the design of future prosthetics. Many veteran amputees, for example, report pain in a variety of locations including the stump, residual limb, neck, and back, along with pain from the missing or “phantom” limb. “For many people,” Resnik said, “this pain is chronic. It was always present, and it was present again a year later at our follow-up. It’s a significant problem.”
The survey also gauged subjects’ interest in advanced surgical techniques such as osseointegration – the attachment of a prosthetic directly to the underlying bone, which avoids a socket altogether and makes an artificial limb a relatively seamless extension of the residual limb – or neurosurgeries that may improve their movement control. “We did find a really strong interest among these veterans with amputation, a willingness to consider surgery to have some features that are now mostly experimental but becoming more available.”
A second study, also aimed at discovering more about how people interact with specific devices, involves in-person data at multiple sites – but in-person data collection sessions were paused due to COVID-19 concerns. “We’ll be resuming it soon,” Resnik said. She’s hoping the study will collect sufficient data from smaller subgroups to focus on more device-specific questions, such as the benefits of having a powered rotator on an upper-limb prosthesis, or a hand with multiple grip patterns.
Resnik’s group is also investigating the needs of woman veterans with upperlimb amputation, who have expressed a slightly different set of needs and preferences – and who have been revealed by the nationwide studies to be less likely to use prosthetic arms than men. “Even though it looked as if, among the people who did use the prosthesis, male and female veterans were about equally satisfied, there were more women who didn’t use a prosthesis,” Resnik said. “So we know that we need to have some sort of way of assessing their perspective on available devices. We’re working on that kind of measure now, and we hope to have something to contribute in that area.”
In collaboration with a team led by principal investigator Dustin Tyler, PhD, of the Cleveland VA and Case Western Reserve University, the Limb Function group has begun work on a study funded by both the VA and the Defense Advanced Research Projects Agency (DARPA) to investigate the impact of a new neural-connected prosthetic. The iSENs system was designed to simulate a sense of touch through an array of fingertip pressure sensors and an aperture sensor, which gives a sense of the robotic hand’s opening or closing. Data from these sensors is transmitted via leads implanted in the user’s arm, which can be used to activate upperarm flexors and/or to open and close the hand. The workings are complex, but essentially this system – an implanted somatosensory neurostimulator system – allows a user to “feel” sensory input and respond by gripping and lifting or moving items.
Resnik was among a VA research team who first reported the qualitative descriptions given by users of the iSENs system: what it felt like; how natural it was; how useful it was for certain tasks; and the degree to which it helped them return to normalcy. She is now a co-investigator for the new VA-funded randomized clinical trial comparing an advanced version of the iSens system to other prostheses. After the post-pandemic resumption of data collection, Resnik hopes it will grow into a multi-site study. “We already have a few people with the initial implanted systems who have demonstrated that it’s possible to create a prosthesis that offers sensation,” she said, “and we have a good sense of the fact that it makes a huge difference to people in terms of feeling like the prosthesis is part of them.” The new system, she said, will offer more sensory abilities as well as improved movement control.
All neural-connected prosthetics are still in the experimental phase, Resnik pointed out; if you’re using one, you’re a research subject. The Limb Function group at the CfNN is working to answer questions about the future: about how interested and willing veterans might be to use these body- or brain-powered devices. But it’s just as important to make sure an upper-limb amputee has the right fit for their lifestyle today.
“One of the things I’ve learned over these years,” said Resnik, “is that one prosthesis doesn’t fit everyone, and that some people benefit from different types for different situations.”
Resnik’s observation – different circumstances require different solutions – is an insight shared by everyone at the VA’s Center for Neurorestoration and Neurotechnology; in fact, it’s one of the center’s founding principles. “Our center brings together researchers from multiple disciplines, including neurology, and neuroscience, and engineering, and physical therapy, and computer science,” said Hochberg. “To bring that group together to inform each other’s research, and to be able to focus that research on improving veterans’ health, is incredibly rewarding for all of us.”