Simultaneous assessment and training of an upper-limb amputee using incremental machine-learning-based myocontrol: a single-case experimental design.
Markus Nowak, Raoul M Bongers, Corry K van der Sluis and 2 others
PMID 37029432WHAT IT FOUND
One transradial amputee learned to control a multi-articulated prosthetic hand using an incremental machine-learning myocontroller.
He achieved reliable performance in daily tasks comparable to his standard prosthesis after 31 sessions, but this single case cannot prove the method works for others.
Key findings
01The participant successfully controlled four distinct actions (power grasp, pointing, preshaping, flat grasp) with the machine-learning system after a failed attempt to add a precision grasp.
02Task completion times and self-rated satisfaction improved within each training phase, eventually reaching a level comparable to the participant's own standard two-sensor prosthesis.
03Incremental updates allowed the user to correct control instabilities in specific positions without needing full retraining sessions.
STILL TO COME
How it was doneWhat they foundWhat it means for PTsWhat it means for OTsWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study involved only one participant, so the results cannot be generalized to other amputees. The participant had 11 years of experience with a standard prosthesis, which may have influenced his ability to learn the new system. The tasks used in the assessment are not yet validated, so it is unclear if they accurately measure real-world functional improvement. Several errors occurred during the study, including incorrect instructions from the experimenter and an unnecessary full retraining, which may have affected the data. One phase of training (adding a precision grasp) failed and was aborted, highlighting that this method does not work for every action combination.
Declared interests
The study was funded by the German Aerospace Center (DLR) under the TACT-HAND project. The authors declared no competing interests.
The easy way to misread this
Do not assume this machine-learning myocontroller will work for your patients. This is a single case report where one experienced user eventually succeeded after a failed phase. The method is not validated, and the results do not prove that this technology is superior or easier to learn than standard prostheses for the general amputee population.