Three people with complete spinal cord injuries regained coordinated leg movement and a functional sense of where their limbs were positioned, according to results published in Nature Biomedical Engineering on March 11. The study, led by researchers at Brown University and Rhode Island Hospital, is the first to demonstrate simultaneous motor and sensory restoration in people with motor-complete paralysis.
The participants, all of whom had lost the use of their legs, received two electrode arrays implanted above and below the site of their spinal cord injury. The arrays below the lesion activated leg muscles to produce stepping movements on a treadmill with physical therapy support. The arrays above the lesion delivered patterned electrical stimulation designed to substitute for the sensory signals that the injury had severed.
How the sensory substitution works
In a healthy spinal cord, proprioceptive neurons constantly relay information about limb position back to the brain. A complete injury severs that pathway entirely. Previous spinal stimulation research, most notably from Grégoire Courtine’s group at EPFL, had demonstrated that continuous stimulation could restore movement but tended to block the remaining proprioceptive signals in humans. Courtine’s team addressed this with burst stimulation protocols that preserved some sensory information. The Brown study goes further by actively generating a substitute sensory channel.
The approach works by mapping stimulation-induced sensations in intact body regions (the chest, arm, or back) to specific joint angles in the legs. Over the course of a two-week in-hospital trial, participants learned to associate these substitute sensations with leg position. When blindfolded, they could accurately report knee joint angles based solely on the stimulation feedback.
David Borton, associate professor of engineering and brain science at Brown and the study’s senior author, trained under Courtine at EPFL before establishing his own lab focused on brain-spinal interfaces. His postdoctoral work included developing the first fully implantable wireless brain sensor in 2013 and demonstrating BCI-driven walking restoration in nonhuman primates in 2016. The current study represents a direct extension of that line of work into human participants.
Machine learning and the DJ board
The research team used deep learning methods to determine the optimal stimulation parameters for each participant. Rather than applying a fixed protocol, the system identified motor- and sensory-specific patterns tailored to each individual’s injury and anatomy.
Participants interacted with a custom interface the researchers called a “DJ board,” fitted with knobs and sliders that controlled which parts of the spinal cord received stimulation, along with the speed and intensity. Lead author Jonathan Calvert, now an assistant professor at UC Davis, noted that participants could explore stimulation patterns in real time. The participants themselves reported that the DJ board was enjoyable to use, which is a nontrivial observation in a trial that required intensive daily sessions.
What this means for the field
Spinal cord stimulation for motor restoration has been demonstrated before. Courtine’s group at EPFL and others have shown that epidural stimulation can enable stepping and even independent walking in some incomplete injuries. What distinguishes the Brown study is the combination: movement and meaningful sensory feedback operating simultaneously in people with complete injuries where no voluntary motor function remained.
The sensory component matters because walking without proprioception is inherently unstable. Even if muscles can be activated, the absence of feedback about foot placement, joint angle, and ground contact makes functional ambulation difficult and potentially dangerous. The fact that participants could learn to interpret substitute sensory signals within a two-week trial window suggests the brain retains substantial plasticity for integrating new feedback channels, even years after injury.
No device-related adverse events were reported across the three participants. The team plans to recruit additional participants for a longer-term study that would test the system outside the hospital setting, a step that would bring it closer to practical, everyday use.
The study was conducted through Brown’s Carney Institute for Brain Science and the VA Providence Center for Neurorestoration and Neurotechnology, with neurosurgical support from Dr. Jared Fridley, now chief of spinal neurosurgery at the University of Texas at Austin.