Mind Over Matter

Video by 7 Wonders Cinema / Multimedia editor: Carly Williams

Before the trial began, Aaron set a goal for himself of riding 350 miles on his adult trike. The day before surgery, he looked at his tricycle’s odometer: 353 miles.

Although intensive rehabilitation over two months helped Aaron regain many abilities – he was able to walk short distances with a cane, to speak and swallow – he, like so many other stroke survivors,  lost his ability to work and a good deal of his independence. He didn’t know it then, but in two years Aaron’s life would change dramatically again. He would become the first person with this common type of stroke to be implanted with brain electrodes – a brain computer interface (BCI) — that would send signals from the brain to a robotic arm brace designed to move his weaker left arm.

But before that future can become reality, Cortimo researchers would have to test whether it’s possible to find usable brain signals still capable of controlling movement in someone like Aaron. It was an open question — one that Aaron and Holly were committed to answering, through two brain surgeries, three months of intensive therapy and much more.

BCI implants were first pioneered in humans some 15 years ago. Patients paralyzed from the neck down were able to use their thoughts alone – as read by these brain-implanted electrodes and interpreted by a computer – to power electrical devices such as a computer cursor or a robotic arm or chair. These patients’ spinal injuries had broken the connections between their brains and bodies, but their brains were otherwise healthy. In stroke patients, however, parts of the brain go dark and die. Could brain implants connect with still living neurons near the area impacted by stroke and could they learn to control movement? Would Aaron’s brain be able to control a fitted, robotic brace to help him improve his fine-motor control of his weakened arm? The Cortimo trial aimed to find out.

This science sketch helps explain how this research is using brain-computer interface (BCI) to restore arm movement after a stroke.

Science sketch by Karuna Meda, PhD

Conquering the Beast

After Aaron’s stroke in 2019, two months of inpatient rehab helped him walk again, but not very far. “During his first day of rehab, he took a total of 12 steps, and collapsed back down, exhausted,” recalls Holly. He got a wheelchair for use at home, but it was too big to navigate inside the house. Holly dubbed it “the beast.” They’d use the chair for longer trips, like walking their dog Gator around the neighborhood.

They’d start their walks with “the beast” at the end of their driveway so Aaron could walk to it. Each day, Aaron asked his mom to move the wheelchair a little bit further, to fence post, then the stop sign, then down the street. Until one day he walked the loop of their neighborhood – 1.2 miles – without needing the wheelchair once.

Holly and Aaron taking a walk in their neighborhood.

“Aaron’s drive to improve, to succeed, to become more independent, that tenacity was such an important aspect to the success of this trial,” says the lead researcher of the Cortimo trial at Jefferson, Mijail Serruya, MD, PhD, who was part of the team that had implanted the first human with brain implants 15 years ago.

We needed someone who was ready to work really hard on something that would be different than anything he’s ever done before. We were incredibly lucky to find just that person in Aaron and his mother, Holly.”

The pre-screening for the trial was slated to begin in March, 2020, with brain-implantation surgery in April. But the plans came to a screeching halt as Philadelphia began to shut down. Schools closed and all non-essential staff had to stay home to slow the spread of COVID-19. All clinical trials at Jefferson, including the Cortimo trial that Aaron had signed up for, were halted.

Time to Reboot

“Aaron was so worried,” recalls Holly. “That whole summer, he kept saying ‘I hope I don’t lose the trial.’” Holly texted regularly with Erica Jones, the clinical trial coordinator for the Cortimo study, who offered reassurance and information as it came in. By the middle of the summer infection rates had slowed, and the restrictions on clinical trials began to lift. The team re-booted their plans. Erica made sure Aaron and his mom could get to all of the appointments from their home, nearly an hour away from Jefferson, with all of the necessary COVID-19 precautions in place. Soon, the team had set a new day for surgery – October 23rd.

Despite the difficulty ahead, despite knowing that he’d have brain surgery twice – once to implant the brain electrodes, and once to remove them three months later – Aaron had become excited by the possibilities that the clinical trial offered him. He’d have the chance for additional intensive rehabilitation through physical and occupational therapy and learn to use a robotic brace fitted for his weakened arm.

The brace, custom-made for Aaron, would be his to keep post trial. More importantly, his participation would inform future treatment for stroke patients like him. “Aaron knew this trial stood to help lots of other people with stroke one day. That really mattered to him,” says Holly.

The Road to Brain Surgery

“His hand is really tight,” says Dr. Serruya at one of the Aaron’s visits with Jefferson rehab occupational therapist Joe Kardine. Using a bit of force, Dr. Serruya gently stretched Aaron’s curled fingers out straight. As soon as he let go, the fingers curled inward again to their resting position. “I’m worried that if we can’t loosen his grip, we’re not going to be able to get the brace over Aaron’s hand,” says Dr. Serruya. “We can’t do the trial without the brace.”

 Occupational therapist Joe Kardine working with Aaron to relax his left hand’s grip.

For a month, Aaron came into the city from southern New Jersey for two sessions weekly with Kardine and physical therapists Tiffany Prince-Kandrakota and later, Joseph McCoy, to reduce the tension in this left hand and to improve his walking. Using electrical stimulators placed on top of Aaron’s arm to send small pulses into his muscles to activate then release them , Kardine worked to loosen Aaron’s grip. After a few weeks, the stiffness had improved and Dr. Serruya noticed less resistance in opening Aaron’s hand.

Aaron with the electrical stimulators on his arm to help open his hand.

While Aaron got ready, so did many other experts working on the Cortimo Project. Three weeks before surgery, Aaron came in for an fMRI with Kiran Talekar, MD, and his radiology team, who were tasked with finding the exact location where the brain implants should go. Although the stroke had damaged part of Aaron’s brain, the team was searching for areas of brain tissue that might still have enough signal to control his left arm via electrodes. Aaron lay in the MRI tube with Dr. Talekar by his side instructing him on when to think about opening his hand, in time with the machine’s scan cycle. Dr. Serruya, physicist Devon Middleton, PhD, and MRI technologist Jamie Noonan sat in the observation gallery, watching the brain scans — dark for many cycles — suddenly light up with activity. They found areas of brain tissue where electrodes could be implanted.

Neuroradiologist Dr. Kiran Talekar showing Aaron an fMRI of his brain tissue.

With these scans and others, neurosurgeon Chengyuan Wu, MD, developed the specifications for a 3D printed replica of Aaron’s brain, created by Jefferson’s Health Design Lab. The scan and the 3D replica would help neurosurgery chair Dr. Robert Rosenwasser, president of the Vickie and Jack Farber Institute for Neuroscience, together neurosurgeons Dr. Wu and Dr. Ashwini Sharan plot out and prepare for the surgery.

 Dr. Rosenwasser with Aaron’s 3D printed brain.

Day Zero

The day of the surgery finally came before sunrise, the morning of Friday, October 23rd. Aaron was excited by the opportunities the surgery might present.

Neurosurgeons Drs. Ashwini Sharan, Robert Rosenwasser, and Chengyuan Wu during Aaron’s procedure.

Holly was a few floors away nervously waiting for updates from Dr. Serruya and the team. Erica Jones had taken a day off work to be with Holly, taking her for walks outside when worry made it too hard to sit still. “He’s grown up. But he’s still my little boy,” Holly says.

The four electrode arrays were successfully placed, with wires leading through Aaron’s skull into two ports externally visible at the top of his head. His three-month, intensive journey to answer the clinical trial question had begun.

Mind Games

Every day, Aaron, Dr. Serruya and his team of engineers, would meet at the residential hotel suite near Jefferson where Aaron stayed. The day would begin with Aaron’s arm under an electrical warming blanket, to release that abnormal tension and relax the curled fingers of his weaker left arm. Then, Dr. Serruya would plug one of the ports sticking out of Aaron’s head into a massive computer that would record the signals from individual neurons just next to his stroke.

Dr. Serruya working with Aaron to control the arm brace.

Now the real work had begun. “This is like learning to walk, but learning it every day, while the floor shifts under you, and the laws of gravity change,” says Dr. Serruya.

Each day, before Aaron could learn to control the robotic brace, the computer algorithms had to learn from Aaron—using artificial intelligence (AI). The electrodes recorded the “voices” of Aaron’s neurons in his brain, as Aaron controlled the movement of a small dot or characters in a simple computer game – with his thoughts. Then, engineer-neuroscientist Alessandro Napoli, PhD, would pinpoint the individual neuronal cells in Aaron’s brain with the strongest and clearest voice. He was looking for cells that “fired” or activated in time with Aaron’s intention to move – a task made infinitely faster with the help of machine learning approaches.

Aaron working with Dr. Napoli on brain activity.

The Breakthrough

Once the researchers had pinpointed the right neurons and trained the computer algorithm, it was time for Aaron to work with his very own engine-red motorized arm brace — first using his residual arm muscle to work the brace and then finally, bypassing the muscle to control the brace with only his brain.

“When we bypassed the brace’s muscle control with the brain control coming from the implants, it worked,” says Dr. Serruya “Aaron said it felt natural. That day was a breakthrough.”

Aaron successfully drops an eraser into a bucket using his arm brace.

There were other times, though that it seemed as if the neurons the research team chose to be the day’s main “voices” weren’t the right ones and Aaron struggled to control the arm smoothly. “If the brace does not respond quickly enough, he fights it,” recalls Dr. Napoli after a difficult week. “As soon as you put the brace on, it frustrates him when it doesn’t always do what he wants.”

What Drs. Napoli and Serruya soon discovered was that the brain signals they were recording in Aaron, a stroke patient, were quite different than what they were expecting based on years of research in patients with spinal cord injuries. They had to change the algorithm design to give Aaron better and more consistent control.

“What we’re studying is much more relevant to all of the people who live with disability from stroke. But for that reason, it’s also more challenging,” says Dr. Serruya.

With wireless implants on the horizon, and with what Dr. Serruya and team have learned from Aaron, it’s conceivable that in the next decade, patients like Aaron could get fully implanted and wireless devices (rather than Aaron’s, which still required a wired connection). They would train hard with an AI specialist and team of occupational and physical therapists, and then go home with a more natural control of a motorized brace that restores mobility.

“For a group of patients like Aaron, who really don’t have a lot of treatment options a year or two after their stroke, this approach could be a game changer,” says stroke neurologist Diana Tzeng, MD, who worked with Aaron to try to pinpoint the cause of his stroke and ensure it didn’t happen again.

The What Ifs

A different possibility – though the Cortimo trial wasn’t designed to test it directly – is even more intriguing. Could an implant, coupled with intensive training and rehab, help retrain a portion of a stroke patient’s brain to form new connections needed to improve mobility? Could there be more sustained improvements in mobility even after implants are removed?

One afternoon, near the end of the trial, after Aaron’s arm had been resting under a heated blanket to relax his muscle tension, Dr. Serruya stood a few feet away. “Let me see your arm,” he said to Aaron. For the first time in nearly two years, Aaron instinctively lifted his hand and extended his fingers toward Dr. Serruya, completely unaided. Dr. Serruya looked at Holly, who was sitting nearby. “I was tearing up,” she said. “I haven’t seen those fingers extend like that in two years.” Not quite believing what they saw, Dr. Serruya asked Aaron to do it again. And he did, just like that, as if his brain was finding a way around the abnormal muscle contraction, to control his arm again without the brace.

Occupational therapists like Kardine say that there’s always more that stroke patients can get back with intensive training. “Although most people’s improvements plateau after about six months, it’s not impossible for additional advancements, especially with the intensive training that Aaron’s been getting,” says Kardine.

“Future studies would need to explore that more definitively,” says Dr. Serruya. “What we observed lasted a few moments, but what if it could be patterned into the brain and last longer? We have a lot more to explore. And we couldn’t have gotten here without Aaron’s altruism, his persistence and commitment to helping others.”

Holly and Aaron at home.