Ismael Seáñez, assistant professor of biomedical engineering and of electrical & systems engineering in McKelvey Engineering and of neurosurgery at WashU Medicine, and Rodolfo Keesey, a doctoral student in his lab (standing), took an in-depth look at how well high-frequency waveforms, or kilohertz-frequency spinal cord stimulation, actually target the neural structures that lead to recovery.
Photo Credit: Rod Keesey
Scientific Frontline: Extended "At a Glance" Summary: Transcutaneous Spinal Cord Stimulation (tSCS) Waveforms
The Core Concept: Transcutaneous spinal cord stimulation (tSCS) utilizes non-invasive electrical waveforms to help patients recover motor function following a spinal cord injury. Recent research evaluates whether newer, kilohertz-frequency waveforms are as effective as conventional, longer-duration waveforms at targeting the neural structures necessary for true rehabilitation.
Key Distinction/Mechanism: Conventional tSCS promotes recovery by recruiting sensory (afferent) nerves, which subsequently activate motor nerves, enabling voluntary movement control and preventing rapid muscle fatigue. Conversely, high-frequency kilohertz waveforms demonstrate poor specificity, bypassing sensory pathways to directly activate motor (efferent) nerves. This direct motor activation requires higher stimulation intensities and severely limits the neuroplasticity required for long-term recovery.
Major Frameworks/Components:
- Sensory Pathway Activation: The optimal rehabilitative mechanism that utilizes existing spinal circuits and brain connectivity to facilitate voluntary motor recovery.
- Direct Motor Activation: The preferential target of high-frequency waveforms, which leads to rapid muscle fatigue and lacks a rehabilitative mechanism.
- Waveform Selectivity: The critical ability of a non-invasive electrical current to penetrate the skin and selectively target specific neural structures.
- Dual-Methodology Testing: The utilization of both human in-vivo experiments and computational models targeting the cervical and lumbar spinal segments to validate neural recruitment differences.
Branch of Science: Biomedical Engineering, Neuroscience, Neurophysiology, and Rehabilitation Medicine.
Future Application: This data will guide clinical best practices, medical device procurement, and the engineering of future neuro-prosthetics, steering the medical community toward highly selective conventional waveforms over higher-frequency alternatives for non-invasive therapies.
Why It Matters: As the first wave of tSCS devices clears FDA approval for motor recovery, this research proves that highly expensive high-frequency devices may be fundamentally inferior for rehabilitation compared to affordable, commercially available conventional waveform alternatives.
Spinal cord stimulation has been used alongside physical therapy to help patients improve movement after spinal cord injury. More recently, using high-frequency waveforms in noninvasive stimulation has become popular to reduce discomfort during spinal cord stimulation, but incorporating this technology can dramatically increase costs and reduce access.
New research from a team of scientists in the McKelvey School of Engineering at Washington University in St. Louis, in collaboration with researchers at the Medical University of Vienna, Austria, and the Friedrich-Alexander University in Erlangen-Nürnberg, Germany, suggests the need for a closer look at this practice to determine whether high-frequency waveforms are actually as effective as the longer-duration waveforms already commonly available.
Ismael Seáñez, assistant professor of biomedical engineering and of electrical and systems engineering at McKelvey Engineering and of neurosurgery at WashU Medicine, and Rodolfo Keesey, a doctoral student in his lab, examined whether high-frequency waveforms actually target the neural structures that lead to recovery as effectively as existing waveforms available for conventional spinal cord stimulation. Through three experiments, Seáñez’s team looked at the mechanisms behind responses prompted by high-frequency stimulation in both a human model and computational models to determine any differences. Results of the research were published May 12, 2026, in Nature Biomedical Engineering.
“Computational models and human experiments revealed that conventional spinal cord stimulation promotes motor recovery by recruiting sensory nerves, which subsequently activate motor nerves,” Seáñez said. “This is counterintuitive, because our end goal is to recruit motor nerves and help the muscles move. However, sensory activation allows patients to voluntarily control the movement evoked by stimulation by allowing other circuits to modulate the synapse and then determine how big or how small the effect is.”
“Also, when you recruit motor nerves directly, initially you get a very nice response, but then the muscle starts fatiguing because you recruit all the neurons in that muscle,” he said. “Sensory activation avoids this. We wanted to see whether high-frequency waveforms also recruit motor neurons through the sensory pathway. If they do not, they could be less effective at improving recovery.”
For this study, Keesey conducted three experiments. In the first, he stimulated nerves in the legs of 28 unimpaired participants to artificially activate the sensory pathway to raise the resting potential of the motor neurons.
“Neurons that fire together, wire together,” Keesey said. “Because we have all these rich sources of connections with sensory activation, we can help with rehabilitation by activating a pathway that interacts with the brain and with circuits in the spine. And that allows us to have this rehabilitative mechanism. However, kilohertz-frequency waveforms were found to be less effective than conventional waveforms at activating the sensory fiber pathway. Therefore, we may miss out on all these connections. This is a bad sign for recovery.”
In the second set of experiments, Keesey and his collaborators recreated the setup in a computational model. In the final set of experiments, they tested spinal cord stimulation of the cervical and lumbar portions of the spinal cord controlling the arms and legs.
“What we see is that the kilohertz-frequency currents require much higher stimulation intensities to elicit muscle responses,” Seáñez said. “We determined that people can tolerate higher intensities, but you need higher intensities to generate a response. Importantly, we find that the high-frequency waveforms are preferentially recruiting the motor efferent fibers rather than the sensory fibers. This suggests that kilohertz-frequency waveforms have poor specificity for the target neural structure of spinal cord stimulation.”
Keesey said sensory fiber recruitment is important because without it, the stimulation simply activates the muscle with less opportunity for rehabilitation.
“You lose these rich sources of input from the brain, from the spinal cord, and all these mechanisms for rehabilitation and the fatigue-resistant prosthetic effect,” Keesey said. “With noninvasive spinal cord stimulation, waveform selectivity is really important. When you can implant an electrode and you want to get this sensory pathway, it’s not difficult: we literally stick an electrode on it and activate it. But if I have to go through all your skin, I need to rely on my waveform being selective for that pathway. And what we found is that the high-frequency waveforms are not selective for that pathway. They are more tuned to activate that motor path. This matters a lot right now. The first wave of tSCS devices is now clearing FDA approval for motor recovery following spinal cord injury. These devices all use high-frequency modulated waveforms, but they can be dramatically more expensive. Not only do we already have commercially available devices that are much more affordable and can deliver conventional waveforms, our research indicates that their waveforms may be more effective for rehabilitation.”
Funding: Funding for this research was provided by the National Institutes of Health’s Eunice Kennedy Shriver National Institute of Child Health and Human Development (K12HD073945) and National Institute of Neurological Disorders and Stroke (K01NS127936); the McDonnell Center for Systems Neuroscience at Washington University in St. Louis; the German Federal Ministry of Education and Research; and the Austrian Science Fund.
Published in journal: Nature Biomedical Engineering
Authors: Rodolfo Keesey, Ursula Hofstoetter, Zhaoshun Hu, Lorenzo Lombardi, Rachel Hawthorn, Noah Bryson, Abdallah Alashqar, Andreas Rowald, Karen Minassian, and Ismael Seáñez
Source/Credit: McKelvey School of Engineering | Beth Miller
Reference Number: bmed051226_01
.png)
.jpg)