“Touching hands, reaching out, touching me, touching you.”
The lead up to the chorus of Neil Diamond’s “Sweet Caroline” has become a ritual in stadiums and bars across America, celebrating one of the most universal human experiences: reaching out to touch another person.
For people living with paralysis after a spinal cord injury, that experience often disappears. Damage to the spinal cord not only interrupts movement but also severs the flow of sensory information from much of the body, leaving many unable to feel touch below the site of their injury.
While brain-computer interfaces (BCIs) already restore a measure of independence by translating thoughts into movement, allowing people to control robotic limbs, computers, and other assistive devices, they have long lacked something fundamental: touch. Now researchers report the strongest evidence yet that restoring artificial touch can be done safely over the long term.
In a study published in Science Translational Medicine, scientists followed five people with cervical spinal cord injuries who received tiny electrical pulses directly to the brain’s somatosensory cortex for periods ranging from nearly three years to a decade. Across more than 173 million stimulation pulses delivered over a combined 27 years of implanted use, the researchers found no serious stimulation-related adverse events while consistently restoring sensations of touch to participants’ hands.
One participant in the study described the difference after researchers restored an artificial sense of touch by stimulating his brain.
Instead of merely watching a hand grasp another person’s hand, he said, it felt as though he were shaking it himself. “It’s the difference between ‘I told a robotic arm to shake your hand’ and ‘I am shaking your hand,’” Charles Greenspon, PhD, assistant professor of neurological surgery at the University of Chicago and the study’s lead author, told Inside Precision Medicine. “That degree of embodiment is completely impossible without sensory feedback because you just don’t believe it’s yours.”
Creating touch
That sense of ownership—of a prosthetic becoming part of the body—is one of the biggest remaining challenges in neurotechnology. Companies including Neuralink, Precision Neuroscience, and Paradromics have largely focused on reading signals from the brain. Today’s most advanced BCIs excel at reading neural activity, decoding a person’s intentions to move a robotic arm or computer cursor. But they are largely one-way systems, sending nothing back to the brain.
Without sensory feedback, users must rely almost entirely on vision, making movements slower, less natural, and more mentally demanding. Restoring sensation by writing information back into the nervous system has remained a much greater challenge—not because the concept is new, but because evidence that repeated brain stimulation is safe over many years has been limited.
Although involving only five participants, the study is the largest and longest safety evaluation of intracortical microstimulation in humans. The findings address a major obstacle to the next generation of BCIs: proving that the brain can safely receive artificial sensory information over many years. If confirmed in larger clinical trials, the work could accelerate the development of truly bidirectional BCIs that both decode a person’s intentions and restore the sensory feedback that makes movement feel natural.
The researchers’ approach bypasses the damaged spinal cord entirely. Participants received two tiny microelectrode arrays implanted in the region of the somatosensory cortex responsible for hand sensation. Brief electrical pulses delivered through individual electrodes produced sensations that participants perceived as coming from specific locations on their hands.
“Once that device is in, we deliver tiny electrical pulses to the brain directly, and that creates the sensations,” Greenspon said. “The brain cannot functionally tell the difference between something that occurs at your hand and travels up your arm to the brain or directly stimulates the brain.”
Built to last
The researchers also wanted to know whether the implants would continue working after years of repeated use. Like any implanted medical device, microelectrode arrays gradually degrade.
The results suggest repeated stimulation does not accelerate that process. Detection thresholds increased only gradually, and even after a decade, about 60% of the electrodes in one participant could still reliably evoke touch sensations, while about two-thirds remained functional throughout the study. Researchers also found no relationship between how heavily individual electrodes were used and how quickly they deteriorated. “No company is claiming that these implants last forever,” Greenspon said. “It’s the same as a knee replacement.”
The study’s primary goal was straightforward: determine whether years of repeated stimulation would trigger seizures, damage brain tissue, or accelerate deterioration of the implanted electrodes. Across 623 hours of stimulation, participants experienced no seizures and no serious adverse events related to the stimulation itself.
Researchers also found no evidence that the electrical pulses damaged participants’ remaining natural sensation. The only stimulation-related side effects consisted of 53 brief episodes of lingering tingling or buzzing after stimulation ended, most resolving within seconds. “It’s one of the few times where you want a null result,” Greenspon said. “We just wanted to say, ‘Look, we didn’t find anything bad.’”
That absence of problems may ultimately be the study’s most important finding, providing the evidence regulators and device developers need before sensory feedback can become a standard feature of future BCIs.
Closing the brain’s feedback loop
For people who may eventually rely on BCIs outside the laboratory, restoring touch could prove just as important as restoring movement. Without sensory feedback, every movement requires constant visual attention. Greenspon compares the experience to leaving the dentist after a local anesthetic: “You can speak perfectly… It’s because you don’t know where your tongue is.”
Artificial touch changes something more subtle as well. Rather than operating a machine, users begin to experience a prosthetic as part of themselves. Researchers believe that sense of embodiment could make future prosthetic limbs not only more dexterous but also less mentally exhausting to use.
The next major challenge is restoring proprioception—the internal sense that tells us where our limbs are without looking. Current sensory BCIs cannot reproduce that ability. Researchers also hope future systems will deliver richer sensory information, allowing users to distinguish whether an object is soft or hard, rough or smooth, or beginning to slip from their grasp.
The findings arrive as brain-computer interfaces transition from academic laboratories to commercial development. Before sensory neuroprostheses become routine treatments, researchers will still need to conduct larger clinical studies.
For decades, BCI research has focused on teaching machines to understand the brain. The next generation may depend just as much on teaching the brain to understand the machine. If that happens, the most important advance may not be giving people the ability to control a prosthetic hand. It may be giving them the feeling that the hand is their own.
That is what made the participant’s handshake remarkable. The breakthrough wasn’t that a hand closed around another person’s. It was that, for the first time, it felt like their own handshake.
The post Sweet Caroline: BCIs That Restore Sensation of Touch Show Long-Term Safety appeared first on Inside Precision Medicine.