The effect of mirror therapy can be improved by simultaneous robotic assistance.
Featured on Physle Daily · Monday, 5 October 2026
After stroke or brain injury, severe arm weakness is notoriously hard to treat, and mirror therapy, where patients watch a reflection of their unaffected arm to trick the brain into engaging the paralysed side, is a well-established option. This trial asked whether adding a robotic glove that physically moves the affected hand alongside the mirror illusion could improve on standard mirror therapy alone. Twenty-four patients with severe hand paresis completed fifteen half-hour sessions over five weeks, split between robot-assisted mirror therapy and conventional mirror therapy. The robot-assisted group showed a significantly greater improvement in a standard arm-function score than the mirror-only group.
Other measures, including a muscle strength index and sensation tests, showed no significant difference between groups, and pain levels stayed stable in both. The main limitation is the very small sample, just 24 completers, which means these encouraging results need replication before firm conclusions can be drawn about clinical practice.
If you or someone you know has severe arm weakness after a stroke or brain injury, you may have heard of mirror therapy, where watching a reflection of the working arm can help the brain reconnect with the weaker side. This study tested whether adding a robotic glove, which gently moves the weak hand at the same time as the mirror trick, works better than mirror therapy on its own. Twenty-four patients took part, with half receiving the robot-assisted version and half getting standard mirror therapy, over five weeks of short daily sessions.
The group using the robotic glove showed a significantly bigger improvement in arm movement scores. Other measures, like grip strength and sensation, didn't show a clear difference between the two approaches, and pain stayed stable for everyone involved.
This is a small study, so while the results are promising, more research with larger groups of patients would be needed before this approach becomes standard care.
This is a summary of research, not medical advice. Talk to your own healthcare provider about anything affecting your care.
This trial addresses a specific clinical question: for patients with severe arm paresis (Fugl-Meyer hand/finger subscore under 4) after stroke or traumatic brain injury, does adding robotic co-movement to mirror therapy improve motor outcomes beyond standard mirror therapy? In this randomised, parallel-group design, 24 patients who completed the full 15-session, five-week protocol were compared on the Fugl-Meyer Assessment upper extremity motor score as the primary outcome. The robot-assisted mirror therapy group showed a significantly greater improvement in this primary motor outcome than the mirror-therapy-only group (p=0.006). Secondary outcomes, the Motricity Index and the FMA-UE sensation subscores (surface sensibility and position sense), showed no significant between-group differences, which should temper any inference that robotic assistance broadly enhances sensorimotor recovery rather than motor function specifically.
Pain remained stable in both arms of the trial, and no adverse effects were reported; qualitative feedback suggested the robotic training was acceptable to patients and therapists, though this speaks to feasibility and tolerability rather than efficacy. The principal limitation is sample size: with only 24 completers, this remains a small trial, and while the primary result is statistically significant, the evidence base for broader adoption or for the non-significant secondary findings is not yet established.
In this small randomised trial of 24 completers with severe arm paresis, robot-assisted mirror therapy produced a significantly greater improvement in the Fugl-Meyer upper extremity motor score than conventional mirror therapy alone (p=0.006).
Secondary outcomes, including the Motricity Index and FMA-UE sensation subscores (surface sensibility and position sense), showed no significant differences between groups, and pain levels remained stable in both.
With only 24 patients completing the study, this is a small trial, and no adverse effects were reported, so these encouraging results require replication before broader conclusions can be drawn.
Source
Restorative neurology and neuroscience
Schrader M, Sterr A, Kettlitz R et al. · 2022
doi: 10.3233/RNN-221263