Transradial TMR improved grasp performance in myoelectric prosthesis users
Multifunctional prosthetic hands have become more capable on paper than they are in daily life, especially for people with transradial amputation who depend on stable, repeatable muscle signals to drive dexterous hand movements. In this PLoS One study, eight participants trained on a multi-articulating prosthesis under myoelectric pattern recognition control before and after targeted muscle reinnervation (TMR), and several measures of hand function improved after surgery. The work addresses a basic clinical problem: a prosthetic hand cannot perform well if the user cannot generate distinct, reliable commands quickly enough to control grip changes, object handling, and release.
TMR is designed to redirect amputated nerves into new muscle targets in the residual limb, creating additional biologic sites that can carry the motor intent once destined for the missing hand. In upper-limb prosthetics, that can matter as much as hardware design. A hand with more joints and more grasp patterns still depends on the quality of the signal stream coming from the person wearing it. When those signals are noisy, the device may be technically advanced but functionally frustrating.
In this cohort, participants completed home use trials before and after TMR surgery, and upper-limb outcome measures were collected after each trial. Three functional measures changed in the favorable direction 9 to 12 months after surgery: the Southampton Hand Assessment Procedure, the Jebsen-Taylor Hand Function Test, and the Box and Blocks Test. Those gains suggest a real-world improvement in the ability to manage objects, coordinate grasps, and use the prosthesis in more controlled and efficient ways. A myoelectric arm system does not help if each command is tentative or inconsistent; after TMR, the hand appears to have become easier to drive.
One of the most clinically important findings came from the offline electromyography analysis. Grip classification error decreased after TMR compared with pre-surgical measurements, indicating more reliable signal discrimination for prosthetic control. In practical terms, that means the pattern-recognition system had fewer chances to confuse one intended grasp with another. For users of multi-articulating hands, less classification error can translate into fewer failed attempts, less cognitive burden, and a smoother sequence of daily tasks such as picking up utensils, handling containers, or reaching for fragile objects.
The benefit was not uniform across every endpoint. The Assessment of Capacity for Myoelectric Control and Activities Measure for Upper Limb Amputees did not differ before and after surgery. That pattern suggests a selective effect rather than a sweeping transformation of every self-reported or performance-based metric. The prosthesis became easier to control in key task domains, yet some broader measures of perceived capacity and activity did not move. In a small study, that kind of split response is clinically important because it points to a specific mechanism: TMR may sharpen the signal used for hand control without automatically changing every dimension of upper-limb function.
The study design also provides a realistic window into how TMR may exert its effect. Several participants reported qualitative improvements in residual limb and phantom limb sensations after surgery. Those sensory changes may help explain why muscle contractions became more repeatable. Pain, uncomfortable phantom sensations, and unstable motor recruitment can interfere with the effort needed to produce consistent myoelectric signals. If TMR reduces that interference, the nervous system may generate cleaner, more reproducible activation patterns for the prosthetic controller to interpret.
Why signal quality matters for dexterous prosthetic hands
Pattern recognition control depends on the prosthesis reading distinct electromyography signatures from the residual limb and matching them to intended movements. For a user, the difference between success and failure can be as small as a contraction that is too weak, too variable, or too easily misclassified. TMR offers a biologically grounded way to improve the quality of those signals by giving amputated nerves new targets and, in effect, additional control sites.
The study’s findings fit that logic. Better scores on hand-function tests and lower grip classification error both point toward a prosthesis that is more responsive to user intent. The improvement emerged months after surgery, which is consistent with the time needed for reinnervation, adaptation, and retraining. The surgical intervention is not merely a structural change in the limb; it reorganizes the interface between the nervous system and the machine.
At the bedside, that has direct implications for counseling. Patients considering transradial TMR are not being offered a guaranteed upgrade in every amputee-specific scale. They are being offered a chance to improve the fidelity of control, particularly for dexterous hands that demand fine command resolution. For some users, that may be the difference between a prosthesis that is occasionally worn and one that becomes part of the daily routine.
The study remains small, with only eight participants, so the results should be interpreted as early clinical evidence rather than definitive proof. Even so, the direction of change is coherent across objective testing and signal analysis. When pain and phantom sensations ease and motor signals become more repeatable, prosthetic grasp control can improve in a way that matters at the level of daily function. The unanswered questions now are practical ones: which patients benefit most, how durable the gains remain beyond the study window, and how much the response depends on the specific prosthetic system paired with TMR.