Swiss researchers use a wireless BCI to help a spinal injury patient walk more naturally
The "digital bridge" combines a brain-computer interface with an implanted pulse generator.
Ever year, more than a million people in North America suffer some form of spinal cord injury (SCI), with an annual cost of more than $7 billion to treat and rehabilitate those patients. The medical community has made incredible gains toward mitigating, if not reversing, the effects of paralysis in the last quarter-century including advances in pharmacology, stem cell technologies, neuromodulation, and external prosthetics. Electrical stimulation of the spinal cord has already shown especially promising results in helping spinal injury patients rehabilitate, improving not just extremity function but spasticity, bladder and blood pressure control as well. Now, in a study published in Nature Tuesday, SCI therapy startup Onward Medical, announced that it has helped improve a formerly-paraplegic man’s walking gait through the use of an implanted brain computer interface (BCI) and novel “digital bridge” that spans the gap where the spine was severed.
We’ve been zapping paraplegic patients’ spines with low-voltage jolts as part of their physical rehabilitation for years in a process known as Functional Electrical Stimulation (FES). Electrodes are placed directly over the nerves they’re intended to incite – externally bypassing their own disrupted neural pathways – and, when activated, cause the nerves underneath to fire and their muscles contract. Researchers have used this method to restore hand and arm motion in some patients, the ability to stand and walk in others and, for a lucky few, exosuits! The resulting limb motions however were decidedly ungraceful, resulting in ponderous arm movements and walking gaits that more resembled shuffles.
Onward’s earlier research into epidural electrical stimulation showed that it was effective at targeting nerves in the lower back that could be used to trigger leg muscles. But the therapy at that time was hampered by the need for wearable motion sensors, and by, “the participants … limited ability to adapt leg movements to changing terrain and volitional demands.“ Onward addressed that issue in Tuesday’s study by incorporating a “digital bridge” to monitor the brain’s command impulses and deliver them, wirelessly and in real-time, to a stimulation pack implanted in the patient’s lower back.
Clinicians have employed these systems for the better part of a decade to assist in improving upper extremity control and function following SCI – Onward’s own ARC EX system is designed to do just that – though this study was the first to apply the same theories to the lower extremities.
Onward’s patient was a 38-year-old man who had suffered an “incomplete cervical (C5/C6) spinal cord injury” a decade before and who had undergone a five-month neurorehabilitation program with “targeted epidural electrical stimulation of the spinal cord” in 2017. “This program enabled him to regain the ability to step with the help of a front-wheel walker,” the research team noted in the Nature study. “Despite continued use of the stimulation at home, for approximately three years, he had reached a neurological recovery plateau.”
In addition to the EX, Onward Medical has also developed an internally mounted electrostimulation therapy, the ARC IM. Per the company, it is”purpose-built for placement along the spinal cord to stimulate the dorsal roots,” to help improve SCI patients’ blood pressure regulation. The system used in Tuesday’s study used the ARC IM as a base and married it to a WIMAGINE brain computer interface.
The Onward team had to first install the BCI inside the patient's skull. Technically, it was a pair of 64-lead electrode implants, each mounted in a 50-milimenter circular-shaped titanium case that sits flush with the skull. The WIMAGINE “is less invasive than other options while offering sufficient resolution to drive walking,” Dave Marver, OnwardMedical CEO, told Engadget via email. “It also has five-year data that demonstrates stability in the clarity of signals produced.”
Two external antennas sit on the scalp, the first providing power to the implants via inductive coupling, the second to shunt the signal to a portable base station for decoding and processing. The processed signal is then beamed wirelessly to the ACTIVA RC implantable pulse generator sitting atop the patient’s lumbar region where 16 more implanted electrodes shock the appropriate nerve clusters to move their legs. Together they form a Brain Spine Interface (BSI) system, per Onward.
The entire setup is designed to be used independently by the patient. The assistive walker houses all the BSI bits and pieces while a tactile feedback interface helps them correctly position the headset and calibrate the predictive algorithm.
In order to get the BCI and pulse generator to work together seamlessly, Onward leveraged a “Aksenova/Markov-switching multilinear algorithm that linked ECoG signals to the control of epidural electrical stimulation parameters,” which seems so obvious in hindsight. Basically, this algorithm predicts two things: the probability that the patient intends to move a specific joint based on the signals it’s monitoring, and both the amplitude and direction of that presumed intended movement. Those predictions are then dumped into an analog controller which translates them into code commands that are, in turn, cycled to the pulse generator every 300 milliseconds. In all, the latency between the patient thinking, “I should walk over there,” and the system decoding those thoughts is just 1.1 seconds.
Calibrating the system to the patient proved an equally quick process. The patient had figured out how to properly “activate” the muscles in their hips to generate enough torque to swing their legs within the first two minutes of trying — and did it with 97 percent accuracy. Over the course of the rehabilitation, the patient managed to achieve control over the movements of each joint in their leg (hip, knee and ankle) with an average accuracy (in that the BSI did what the patient intended) of around 75 percent.
“After only 5 min of calibration, the BSI supported continuous control over the activity of hip flexor muscles,” the team continued, “which enabled the participant to achieve a fivefold increase in muscle activity compared to attempts without the BSI” Unfortunately, those gains were wiped away as soon as the BCI was turned off, instantly losing the ability to step, they explained. “Walking resumed as soon as the BSI was turned back on.”
It wasn’t just that the patient was able to graduate from walking with a front-wheeled frame walker to crutches thanks to this procedure – their walking gait improved significantly as well. “Compared to stimulation alone, the BSI enabled walking with gait features that were markedly closer to those quantified in healthy individuals,” the Onward team wrote. The patient was even able to use the system to cross unpaved terrain while on their crutches, a feat that still routinely proves hazardous for many bipedal robots.
In all, the patient underwent 40 rehab sessions with the BCI – a mix of standard physio-rehab along with BCI-enabled balance, walking and movement exercises. The patient saw moderate gains in their sensory (light touch) scores but a whopping 10-point increase in their WISCI II scores. WISCI II is the Walking Index for Spinal Cord Injury, a 21-point scale measuring a patient’s ambulatory capacity ranging from 20, “can move zero assistance,” down to 0, “bed ridden.“ Onward’s patient went from a 6 to a 16 with the help of this therapy.
“As the participant had previously reached a plateau of recovery after intensive rehabilitation using spinal cord stimulation alone, it is reasonable to assume that the BSI triggered a reorganization of neuronal pathways that was responsible for the additional neurological recovery,” the Onward team wrote. “These results suggest that establishing a continuous link between the brain and spinal cord promotes the reorganization of residual neuronal pathways that link these two regions under normal physiological.”
While the results are promising, much work has yet to be done. The Onward team argues that future iterations will require “miniaturization of the base station, computing unit and unnoticeable antennas,” faster data throughputs, “versatile stimulation parameters, direct wireless control from the wearable computing unit,” and “single low-power integrated circuit embedding a neuromorphic processor with self-calibration capability that autonomously translates cortical activity into updates of stimulation programs.”
Despite the daunting technical challenges, “the BCI system described in Tuesday’s Nature publication may reach the market in five to seven years,” Marver predicted. ”It is possible and realistic that a BCI-augmented spinal cord stimulation therapy will be on the market by the end of the decade.”