A parallel classification strategy to simultaneous control elbow, wrist, and hand movements.
Francesca Leone, Cosimo Gentile, Francesca Cordella and 3 others
PMID 35090512WHAT IT FOUND
A muscle-signal classifier identified simultaneous elbow, wrist, and hand movements more quickly and reliably than the usual classifier in healthy people wearing six sensors.
Offline scores were similar. It was not tested in amputees or with a real prosthesis.
Key findings
01Offline classification scores were not significantly different between the logistic-regression and linear-discriminant classifiers for the elbow, wrist, and hand.
02In real time, the logistic-regression method completed intended motions faster than the linear-discriminant method: mean completion time was 1.84 ± 1.25 seconds versus 2.49 ± 1.87 seconds across the 27 motion classes.
03In real time, the logistic-regression method had higher completion rates than the linear-discriminant method for several combined elbow, wrist, and hand motions, especially motions involving wrist supination.
STILL TO COME
How it was doneWhat they foundWhat it means for OTs
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
All participants were healthy, so the study does not show that people with upper-limb amputation or targeted muscle reinnervation could use this control with a prosthesis. The study included 15 people and real-time results came from two repetitions of each motion class. The tested task was producing gestures while wearing sensors, not using a real prosthetic device. Offline classification scores were not significantly different between methods, so the advantage was mainly in real-time completion measures. Wrist and hand combined movements were more difficult than elbow movements, especially wrist supination.
Declared interests
Funded by Istituto Nazionale per l’Assicurazione Contro Gli Infortuni sul Lavoro and H2020 European Research Council. The supplied text does not report author conflict-of-interest declarations.
The easy way to misread this
Do not conclude that this improves prosthetic function for amputees. The study tested only 15 healthy people in a laboratory, not amputees or people using a real prosthesis, and the offline scores were not significantly different from the benchmark method.