Back

A wireless magnetic implant system for continuous neuromuscular sensing

Shallal, C. C.; Taylor, C. R.; Yeon, S. H.; Casler, R. J.; Oseyemi, D.; Herrera-Arcos, G.; Qiao, J.; Shu, T.; Xu, J.; Levine, D.; Rajcevic, A.; Boerhout, S.; Liu, A.; Clarrissimeaux, E. G.; Paradiso, J. A.; Carty, M. J.; Herr, H. M.

2025-11-22 rehabilitation medicine and physical therapy
10.1101/2025.11.17.25340274 medRxiv
Show abstract

Precise measurement of neuromuscular mechanics yields an intuitive control signal for producing synchronous movement with wearable robotics. Practically obtaining such measurements remains challenging as current muscle-sensing technologies excessively compromise between signal fidelity, system complexity, and invasiveness. Here we present a skin-mounted, magnetometer array platform that wirelessly tracks passive 3 mm diameter magnetic beads implanted within human muscle tissue for continuous neuromuscular sensing. The system employs customized high-density sensing electronics with an information-theoretic architecture to achieve sub-millimeter resolution of real-time muscle dynamics at tracking depths of up to 6 cm within the body. We deploy the platform in a first-in-human clinical study to track a constellation of permanently implanted magnets to enable multiple degree-of-freedom neuroprosthetic control. We demonstrate that the wireless muscle state estimation can outperform standard surface and implanted electromyography interfaces to achieve more accurate and responsive neuroprosthetic movement. Further, we successfully supplant electromyography altogether by extending the platform to detect muscle activation through magnetic induction alone. One-Sentence SummaryA skin-mounted magnetometer sensing array can wirelessly track 3 mm diameter permanent magnetic implants in humans to provide precise neuromuscular information for improved neuroprosthetic control compared to electromyography.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.