At Friedrich-Alexander-Universität Erlangen-Nürnberg, researchers studying electrical stimulation of the spinal cord have uncovered a troubling inefficiency: the high-frequency pulses widely used in today's medical devices may be missing the very nerve fibers that matter most for recovery. The finding, published in Nature Biomedical Engineering and arrived at through human electrophysiology experiments and computational modeling, suggests that millions of patients with spinal cord injuries deserve better solutions.
Spinal cord injury is usually irreversible, yet intensive training combined with medical technology support has given chronically paralyzed individuals a genuine chance to relearn movement. Early breakthroughs came from invasive spinal cord stimulators positioned directly near nerve roots—but the procedure itself is traumatic and technologically demanding. A less invasive alternative emerged: electrodes placed on the skin above the spinal cord that deliver current noninvasively. Clinical trials proved the concept worked, and the first commercial devices reached European and American markets.
But the medical community has largely been operating in the dark about why these devices work and how to optimize them. "We have established that there is little well-founded knowledge about why these products work at all and how they should be applied in a targeted manner," explains Prof. Dr. Andreas Rowald of FAU's Chair of Digital Health. That knowledge gap is precisely what his international team—drawing also from the Medical University of Vienna and Washington University in St. Louis—set out to close.
The researchers tested 28 healthy subjects to map exactly which nerve and muscle activations occur during noninvasive electrical stimulation, focusing on the arms, legs, cervical spine, and lumbar spine—the regions most relevant to spinal cord injury. Simultaneously, they ran high-resolution computer simulations using what Rowald calls "digital twins of the human body," sophisticated models incorporating biophysical data that range from macroscopic current flow through tissue down to the microscopic modulation of individual ion channels on neural membranes.
The comparison proved illuminating and sobering. The clinical activation patterns diverged in important ways from what conventional wisdom had assumed. Most strikingly, the widely adopted high-frequency, ultrashort pulse approach appeared inefficient—failing to reliably activate those deeper nerve fibers believed to be crucial for therapeutic benefit. The models revealed why: at higher frequencies, the pulses struggle to penetrate and recruit the specific neural populations that drive functional recovery.
The implications are significant. If current commercial devices are missing key activation targets, their clinical results may be limited by design rather than by the underlying biology of spinal cord injury recovery. Patients who could potentially regain more motor function are being treated with suboptimal technology, and clinicians lack the evidence base to refine their approach.
"The models allow insight into processes that cannot be directly observed experimentally in humans," Rowald notes. That capability has positioned FAU's team to predict precisely how different electrical parameters would activate nerves and where electrodes should be positioned for targeted responses. The findings open a pathway not just toward understanding why existing devices work, but toward designing substantially better ones—devices informed by rigorous science rather than inherited assumptions.