University of Tennessee researchers received $809,000 to build an implantable prosthetic hand controller that reads mechanical muscle signals. Grip force feedback is also part of the design.

Up to 45 percent of people who receive a powered prosthetic hand eventually stop using it. That number appears often enough in the research literature that it has become the standard framing for why new work in prosthetic hand control matters. Three researchers at the University of Tennessee, Knoxville are now working on one specific layer of that problem: how the hand reads what the user is asking it to do.

The signal problem in current prosthetic hands

Most powered prosthetic hands available today use myoelectric control. Electrodes placed on the skin surface pick up electrical signals from residual muscles, and the controller interprets those signals to drive the hand. The approach has been in clinical use for decades.

The limitation is signal quality. Electrical signals from muscles are diffuse and variable. They shift depending on electrode placement, skin condition, sweat, and limb position. The signal is real, but it is noisy enough that precise, consistent control across the full range of daily tasks remains difficult for many users.

Dustin Crouch, Jindong Tan, and Nicole McFarlane at UT Knoxville’s Tickle College of Engineering are taking a different approach. Rather than reading electrical signals from skin-surface electrodes, their Myo-Bridge implant would sense mechanical signals: the physical motion of residual muscles as they contract. The implant reads that motion and relays it wirelessly to an external prosthetic hand.

Crouch described the mechanism in a press release issued through EurekAlert!: “There’s an implanted component that’s sensing the motion and forces of the muscles, and that information is communicated wirelessly to the prosthetic hand, so it moves when the muscles are moving.”

The return direction: feeling what the hand is holding

The device is designed to work in both directions. The control signal goes from residual muscles to prosthetic hand. The Myo-Bridge also aims to send information back: when the hand grasps something, data about grip force would travel to the implant and create a mechanical lock. The user’s residual muscles would register that lock as resistance, as though holding the object directly.

Proprioceptive feedback, the sense of what a limb is doing and what it is encountering, is absent from most prosthetic hands in current clinical use. Users can see what the hand is doing but cannot feel through it. The Myo-Bridge approach would not restore biological sensation. It aims to create a mechanical substitute: a signal encoding grip state, routed back to tissue that can register it.

Tan described the design intent: “This Myo-Bridge implant extends the traditional exoskeleton and prosthetics concept toward an ‘endoskeleton,’ a robotic system deeply integrated with living tissues. The goal is to develop bio-symbiotic systems in which the robotic and biological components function together as one integrated system, enabling bidirectional interaction and feedback.”

Funding, testing, and where the project sits

The National Science Foundation awarded the team $809,000 through its Mind, Machine, and Motor Nexus program, which supports research on safe, productive interaction between people and engineered systems. The grant supports two graduate students and includes outreach to high school students as well as engagement with amputees and clinicians to help guide development.

The team’s first step is a physical prototype. Before testing with amputee participants, a wearable version of the device would be tested on people without limb loss, who would wear it externally while controlling a prosthetic hand temporarily attached to their forearm. Those results would be compared against conventional myoelectric control.

This is standard sequencing in device research: demonstrate the concept on a simpler population before moving to trials that require recruiting people with limb loss.

On timeline, Crouch said in the release that the earliest the Myo-Bridge could be widely available would be 8 to 10 years. That assumes a path from prototype through wearable validation, then to trials with amputee participants, then to regulatory review and clinical deployment. Each of those steps has its own failure modes.

What this means

Approximately 15,000 people in the United States experience forearm amputation each year, according to the project team’s grant materials. The powered prosthetic hand abandonment rate suggests that what currently exists does not work well enough for a substantial portion of users. The issue, as the team frames it, is not that the devices are poorly engineered. It is that the control interface does not match what a hand needs to do across the variety of tasks people actually use their hands for.

The Myo-Bridge addresses one layer of that problem: better signal reading from residual muscles and a feedback loop that lets users feel what the hand is holding. If mechanical sensing is more reliable than electrical surface sensing for this purpose, and if the feedback loop creates a usable sense of grip, the combination could change the daily picture for people who have found current devices too imprecise to use. Whether this particular approach survives the decade it would take to reach the clinic is what the next phase of research is meant to find out.