NextFin News - A research consortium led by Chalmers University of Technology has successfully decoded complex leg movements directly from the peripheral nerves of above-knee amputees, marking a definitive shift from mechanical automation to biological intent in prosthetic technology. The study, published this week in Nature Communications, utilized ultra-thin hair-like implants and a novel Spiking Neural Network (SNN) to interpret the "neural code" of the sciatic nerve. For the first time, researchers demonstrated that even in cases of severe limb loss, the nervous system continues to transmit precise motor commands—including the specific intent to wiggle toes—that can be captured and translated into digital action.
The breakthrough addresses a long-standing disparity in bionic research. While upper-limb prostheses have long leveraged residual muscle signals for control, leg prostheses have remained largely "dumb" devices, relying on internal gyroscopes and mechanical sensors to react to the environment rather than the user’s will. Giacomo Valle, assistant professor at Chalmers and a lead author of the study, noted that the challenge lay in the extreme weakness of nerve signals post-amputation. By bypassing residual muscles and tapping directly into the tibial branch of the sciatic nerve, the team achieved a resolution of movement control previously thought impossible for lower-limb patients.
At the heart of this interface is a departure from traditional artificial intelligence. Instead of the continuous numerical processing used by large language models, the researchers employed Spiking Neural Networks. These systems mimic biological neurons by processing discrete electrical impulses, or "spikes," in real-time. This alignment with the body’s own communication protocol allowed the team to extract high-fidelity movement intent from relatively limited data sets. Elisa Donati, a professor at the University of Zurich and ETH Zürich, emphasized that this biomimetic approach is essential for developing low-power, fully implantable systems that do not require the massive computational overhead of standard AI.
The implications for the medical device industry are substantial. Current high-end prosthetic legs, such as those manufactured by Össur or Ottobock, can cost upwards of $50,000 yet still lack the intuitive "feel" of a natural limb. The Chalmers study utilized a bidirectional system, meaning the same four electrodes used to read motor intent can also be used to stimulate the nerve to provide sensory feedback. This "closed-loop" capability—allowing a user to both move a prosthetic foot and feel the ground beneath it through a single interface—could drastically reduce the cognitive load required for walking and improve balance in elderly or high-activity amputees.
While the study was a proof-of-concept involving two participants, the accuracy of the decoded movements suggests a clear path toward commercialization. The researchers were able to map specific nerve signals to knee, ankle, and toe flexions with unprecedented precision. The next phase of development involves integrating this neural decoder into a physical prosthetic leg for long-term home use. Success in this area would not only redefine the standard of care for the millions of people living with lower-limb loss but also establish a new framework for how humans interface with wearable robotics.
Explore more exclusive insights at nextfin.ai.
