OtherJournal of neuroengineering and rehabilitation2021

Learning regularized representations of categorically labelled surface EMG enables simultaneous and proportional myoelectric control.

Alexander E Olsson, Nebojša Malešević, Anders Björkman and 1 others

PMID 33588868

WHAT IT FOUND

In able-bodied users, a new algorithm (MRL) controlled a virtual cursor more smoothly and quickly than the standard method (LDA), achieving higher throughput.

Both methods remained stable after seven days without recalibration. This suggests MRL could improve prosthetic control intuitiveness.

Key findings

01MRL significantly outperformed the standard LDA method across all five performance metrics, including completion time and throughput.

02Neither MRL nor LDA showed significant performance deterioration after seven days of no use, indicating stability without frequent recalibration.

03Users preferred MRL over LDA, describing it as smoother and less jittery, while LDA sometimes produced unpredictable cursor movements.

STILL TO COME

How it was doneWhat they found

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.

Already have one?

What it does not show

The study used able-bodied subjects, not amputees, so results may not directly translate to clinical prosthetic users. The test involved a virtual cursor, not a physical robotic limb. The number of degrees of freedom (DoFs) tested was limited to two (wrist and hand), which may not scale easily to more complex prosthetics. The stability finding over seven days may be influenced by the large pickup area of the armband electrodes, which might not be typical for clinical setups.

Declared interests

The authors declare no competing interests. Funding was provided by Stiftelsen Promobilia, Crafoordska Stiftelsen, the European Commission, Vetenskapsrådet, and Lund University.

The easy way to misread this

Do not conclude that this algorithm is ready for clinical prosthetic use or that it works for amputees. The study was conducted on able-bodied individuals controlling a computer cursor, not a physical prosthetic limb.

Read it on PubMed →


The study

Participants
20 able-bodied subjects
Certainty of evidence
Low

Browse

    Cite

    Alexander E Olsson, Nebojša Malešević, Anders Björkman, et al. Learning regularized representations of categorically labelled surface EMG enables simultaneous and proportional myoelectric control. Journal of neuroengineering and rehabilitation. 2021.

    Read the original