What prosthetic arm to use? [flow chart / algorithm]

Algorithm to chose best option for prosthetic arm, based on eight years of consecutive experience and extensive hard work.

Play, tinker, try-out, teaser tests, get an implant just to  see when it breaks, wear a strange socket to see just how long until your skin turns inside out - all cool as long as one has the play money for it and as long as one does not mistake these for actual orthopedic solutions to actual orthopedic problems (that are: real need for bimanual support, avoiding asymmetry and overuse related problems).

A well built cosmetic arm can work towards balancing the body, but usually its awkward mechanics and ill designed hand shape cause more body / trunk contortion for any given repetitive work than assumed; try typing with an iLimb as opposed to typing with a hook, and do that for 7 days a week with four weeks a month, only then will you know what I mean. Sure I could get a bone implant on my 2 cm radius and 1,3 cm ulna bone, but I have to apply some 10 kg oblique / tangiential control cable force and some > 40 kg push / pull / shear forces routinely when handling work items. I cannot just take off three years to wait until my bone decides to grow after babysitting it into submission over countless hours of practice; just as much as I am not interested in some transplant that will ruin the fun for my immune system. Also you do not want to confuse a standard usual body powered arm that prosthetists may build just to convince you to pay them a lot more for electronic gadgets with a really well built body powered arm that actually works for ages without repairs; I was after specific modifications and improvements of my body powered arm for three years until we got that one right. So that is what one wants to be after when concerned with orthopedic aspects.

You have to accept that from me and work from there. I have 8 years experience with this; my overuse / asymmetry issues are down on a bearable minimum while working full force.


The algorithm is actually not dependent on some other factors. Most noteably, it does not depend on cognitive aspects, as suggested by a recent paper by Laura Hancock et al. (2016)  [1].

This becomes apparent when one studies the actual reasons why that diagram is as it is, and why (also) myoelectric arms are inferior, orthopedically, given the premises (you actually do realize that this blog is titled "right below elbow amputee issues" and you came here on your own behalf I presume?):

  • Socket fit is a major problem with myoelectric and other experimental prosthetic arms. Socket fit does not correlate with cognition, just as much as it correlates with total persistence, a different personality trait. Of course mental power is required to beat the environment into such a degree of submission that one in fact ends up with a great socket fit.
  • Grip mechanics is in no way improved by mental performance - unless we are talking shop works, actually working on the prosthetic hand. However, grip mechanics is the end all be all of getting anywhere with a prosthetic arm. People usually do not understand grip mechanics.
  • Myoelectric controls are not cumbersome because they are mentally hard to grasp, but because they contain an ill design. Their unintended co-activation can simply not be avoided reliably, as long as fast fluid motion and orthopedic symmetry is the goal. They crap out 5 minutes into vacuum cleaning the house just because of sweat. And you do not seriously want to tell me that it is a cognitive achievement of an arm amputee to avoid sweating while working? (of course you would want to, but then, that is one of the secret participant requirements for ALL myoelectric studies - no sweat, no immediate hard work - and not one of the big openly declared super pluses)
  • Complexity of a prosthetic arm as such is a bit of an affront, a problem, an obstacle if all one wants is live a life where the prosthetic monstrosity is NOT at the center of attention all the time. If all one wants is a prosthetic arm that seamlessly improves quality of life manually, and that succeeds mid to long term in reducing if not avoiding asymmetry and overuse problems, then it is a matter of sheer cognitive performance to recognize the better solution (i.e.: body powered) and to also revise the deficient parts (i.e.: cable setup, shoulder anchor, wrist unit).

To no surprise, that study contains one particular deficit in listing study participants' requirements: they omitted the fact that participants did not experience a really well built customized performance boosted body powered arm and worked in a manually demanding profession before volunteering for that study. Because really, that is the head to head race that they would want to perform.

Technically, the authors of that study there  [1] are wrong on some other levels - but that is beyond the scope of this recommendation; suffice to say that I am a member of both Mensa and Triple Nine Society since over a decade (just to get any cognitive performance bias out of the way) and if then I say that a myoelectric arm sucks big time for most sensible applications, and as I am a person that has manual work and hobby activities somewhat beyond the average street person then it can be OK to take my word for it.

More concisely, the study summarizes its key findings as "Cognitive domains of attention and processing speed were statistically significantly related to proficiency of DEKA Arm use and predicted level of proficiency"; really, however, they listed 54 single correlation results of various neurocognitive and manual tests of which 19 are negative correlations, none of which exceed a regression coefficient of 0,6, leave alone 0,9 - and only in 9 out of 54 correlations yielded "statistical significance". Furthermore, excellent verbal neurocognitive tests almost exclusively (8/9 tests) correlated negatively with better prosthetic training outcome in a majority even statistically significantly (6/9).

Which means that correlating neurocognitive and manual tests for that type of arm training results in an overall ambiguous result (35% negative correlations) with rare statistical significance for most aspects (only 16% of cross correlations were "significant")  - with the exception that particularly verbally intelligent participants performed particularly badly on prosthetic training, and that rather consistently.

In plain English, that most likely means, that the researchers sucked at explaining stuff, which, however, becomes also apparent just when I check and re-analyze their data (which, as it appears, they failed at correctly describing and interpreting themselves).

The correct sentence in their abstract should read "Cognitive domains of attention and processing speed resulted in somewhat contradictory trends related to proficiency of DEKA arm use and predicted level of proficiency; the only consistent results that were obtained indicated a flawed verbal instruction process". In fact, I have yet to meet the researcher or prosthetic specialist that adquately describes the function of a myoelectric arm for me so it correctly addresses effective function, advantages and disadvantages. Based on that research one may even conclude that in order to arrive at the conclusion "must - do - research - with - "bionic" - arms", it is a necessary requirement that elements of feeling and thinking may have to be absent. But, as we all must assume, less can be more ; )

If it is not a permanent problem, who knows, their thinking just may be a bit clouded. But it is astonishing where they always get these clouds that cloud all that thinking! Usually, once the actual premises are sorted out and clear thinking is introduced, clouds can go away, and prosthetic arms that function can be defined and built.

Used to obtain flow chart @ code2flow:

you are a right below elbow amputee;
if(hard bimanual work,
or overuse/pain/asymmetry?) {
develop own prosthetic arm parts and
network with those that really do,
preferrably parts within "high" load / strain segment
should be used, typically body powered
setups are worlds better at this, avoid
low torque / heavy components, steer
towards highly modular fast switch and
fast replaceable parts
pain, asymmetry?){improve sports training,
increase prosthetic performance
towards more grip performance and
even better durability
else {if
(is there a real reason to
wear a prosthetic arm
just for looks?) {wear cosmetic arm,
tinker with 3d printing,
try other experimental
stuff as long as you
can get out of it
just as easily as
getting into it,
or try to "tinker"
a "bionic" arm
into submission;sick of it?}{chill
no prosthetic arm} }


[1] L. Hancock, S. Correia, D. Ahern, J. Barredo, and L. Resnik, "Cognitive predictors of skilled performance with an advanced upper limb multifunction prosthesis: a preliminary analysis," Disability and Rehabilitation: Assistive Technology, pp. 1-8, 2016.
  title={Cognitive predictors of skilled performance with an advanced upper limb multifunction prosthesis: a preliminary analysis},
  author={Hancock, Laura and Correia, Stephen and Ahern, David and Barredo, Jennifer and Resnik, Linda},
  journal={Disability and Rehabilitation: Assistive Technology},
  publisher={Taylor \& Francis}

Cite this article:
Wolf Schweitzer: swisswuff.ch - What prosthetic arm to use? [flow chart / algorithm]; published 17/04/2016, 14:11; URL: https://www.swisswuff.ch/tech/?p=5844.

BibTeX: @MISC{schweitzer_wolf_1643210848, author = {Wolf Schweitzer}, title = {{swisswuff.ch - What prosthetic arm to use? [flow chart / algorithm]}}, month = {April}, year = {2016}, url = {https://www.swisswuff.ch/tech/?p=5844} }