Situation

As I currently recover from shoulder surgery (left shoulder, biceps long tendon tenodesis, supraspinatus suture / anchor, bumpy ride) this surgery somewhat puts my left arm to a relative rest for a few weeks. That means, as right below elbow amputee (check with blog title?), … what exactly? Correct: glove/s come/s off.

From fully relying on the device, I find new answers to:

• What is the best prosthetic arm?
• What is the best prosthetic hand?
• What is the best prosthetic gripper?
• Is a myoelectric prosthetic arm or a body-powered arm better?
• What are criteria for finding out which is better?

(C) own artwork/AI

The interesting question here is how (almost) fully relying on the prosthetic arm the spells out for prosthetic arm use. And it does spell out.

Significance

As a first effect, this situation shifts a type of significance to the right arm that wasn’t there before. What are the practical differences that I note in particular?

(C) own artwork/AI

1. Universal use with wide grip – The prosthetic grip has to be more universal, and grip more objects than when using it only as assistant arm / hand. There, I find that larger objects are a particular problem with my usual favorite device, the Hosmer 5 hook. Immediately, terminal devices that allow for larger objects are better. At once. I cannot be coerced to help grasping or lifting an object with my left arm just because the prosthetic device isn’t up for it. All of a sudden that matters.
2. Adjustable grip force – The prosthetic grip now necessarily has to have a lower power option: the shoulder anchor simply hurts, and that pain clearly is force/pressure dependent. So I want to reduce or adapt grip strength to a necessary minimum. Sure I am not made of ivory or clay and can sustain pain, no question – but it helps keep chronic irritation down. So, adjustable power or optional VC (or even VC just as is) are more desirable than ever.
3. No object slip/drop – I accept object slip/drop even less than before. I never liked things to slip, and now I find it unacceptable, non-negotiable. The gripper must provide a tight fit to the object, to any object. Slips/drops usually invoke fast/jerk like corrective action with my left hand, and that is not going on right now.
4. Less tolerance for control reliability issues – Control drop-outs are far less acceptable than they ever were. I never liked control failure and now this is a clear reason to not use a device if it isn’t performing at any realistic use rate. Slips/drops usually invoke fast/jerk like corrective action with my left hand, and that is not going on right now.
5. Good grip angles – Grip angles are far less acceptable if they require contortion artistry. So far I found a bit of torso bending not a problem particularly for occasional or rare occurrences, but now I am less tolerant. I want angles to work without me bending over simply also because that hurts or pains or is at least uncomfortable. A single or occasional awkward grip does not matter at all. Once I have to repeat a task, it really starts to become an issue.

The sports training analogy, if you want one, is that of swim stroke training: there, you will first make sure that you are tired out and drained and in a little bit of pain. That is usually achieved by swimming a series with 200 meter or 400 meter distances and aerobics strain and after that, some sprint series. After that, you can swim quite slowly and feel the muscle strain wherever the stroke is problematic. Then that can inform you of a better stroke. A similar effect is at work here: even with a little bit of things going aside with the prosthetic arm device, I will immediately feel “corrective” discomfort or pain. A physiotherapist also explained to me that pain is to be understood as a type of guide or rail, when navigating post surgical recovery. So, it is not necessarily to be totally avoided at all times but it helps to reduce it, keep it low, while letting it inform all (orthopedically relevant) activities or tasks. As I rely on the prosthesis for far more, I also rely on it for things that must not ever drop. With a recently operated shoulder, my mental tolerance of control failure is zero (0.0000000, not even below 0 below any machine epsilon, no, we are talking zero). That means that the engineers that thought they can sneak in a 10% error device? Their things now stall.

Interestingly, only one developer of the many that I interacted with over the years now said that he regularly wore the prosthetic grippers that he built himself. In his workshop. To work. To do things. He was extremely successful with his resulting design. No other designer ever did that and the difference can be felt, is tangible, goes under the skin. No one cares, I know that too – and while the known ingredients for good mechanical design are known, the secret ingredient for good prosthetic arm design is skin in the game.

Choice of devices

Points to be awarded for each category: 0-5, with 0 being worst and 5 being best. The points are then multiplied to give a realistic use rating, and a different metric (that I call “Vidal Sassoon” metric – “if you don’t look good we don’t look good” – an attitude that a number of academic researchers and industrial developers for prosthetic arm technology still could consider to adopt) where I multiply the lowest awarded point with the sum of all points.

Does the result reflect my subjective overall experience? I performed a subjective rating when using these devices 1-2 weeks after shoulder surgery. My shoulder brace did not allow for much pressure. Maybe I underestimated a usefully wide device opening / grip width when it really just could be opened by pulling a bit more but pressure caused too much pain then. Now that I write this >8 weeks after the surgery, I would rate the devices differently again. We can always agree on different base line standards, obviously. This is for after surgery and with little to no tolerance for device failure aspects. 

Yes. The result reflects my subjective overall experience.

But I was surprised how low the Hosmer hook ranked in comparison. How bad it felt on my operated shoulder that the grip forces couldn’t be easily lowered/adjusted.

V2P Prehensor wins. Myoelectric arms remain junk [link].

The TRS Adult Prehensor – exclusively VC – comes second; it lacks features to use cutlery conveniently. Instead the Toughware Equilux , that also comes second, is perfect there and it is the only combined VC/VO device far and wide. But VC suddenly is a life saver. Didn’t see that coming.

The third rank clearly is the indispensable Hosmer 5 Hook despite its well known short comings – it is still dependable and reliable as ever, although uncomfortable due to inability to adapt grip forces.

The Becker Hand suffers from absence of live grip force change and surprisingly small grip opening – which looks good as a hand shaped device, but which is a practical problem once one starts to do everything with it – all of a sudden a grip width of ?~10 cm feels like a must-have. I only realize now that this is a necessary ingredient. Why industrial engineers make everyday things that wide, I could not say, ask them. At any rate, a Becker Hand is still by far the best device if it has to look like a hand. 

The TRS Jaws fared a bit less well than anticipated too: it needed longer prongs and certainly also ubiquitous non-slip rubber covers; these are modifications I was not inclined to work out now and quite frankly, I did not see that coming either. At least that particular device – the TRS Jaws still is absolutely best for bicycle riding – has great potential and at least it can be modified/adapted.

The same cannot be said about the “dead horse” as the big elephant in the room. An elephant that is hovering over people that are deep into hallucinogenic mushrooms. Sacred rituals may come to mind. The urging question imposes itself whether all this much-ado about myoelectric junk can be understood a lot better from entirely non-rational non-scientific angles. Sacred rituals as a true reason for millions of wasted research money [link] makes a lot of sense if one regards the true social standing of disabled people in society.

(C) own artwork/AI

There must be a sense to that, it is just not easily penetrable.

They said, myoelectric arms are for the superbly intelligent¹, they said. They also said, myoelectric arms aren’t reliable – oh no, the fun to be had is that they are explicitly if not proudly unreliable, they said².

I did try the myoelectric setups but these all failed. They are not only junk, they always were [link]. Saw that coming in a way but I was surprised how little tolerance I had for the perpetual failures of myoelectric control. Training, skin surface conditioning, all are not sufficient to prevent control failure there. I do not forgive control errors when I rely on the prosthesis. These are hard words, not fancy buzz words. I wrote “rely”, right? Anyway, if Zö Pröfessörs put their money together to actually pay for my extensive damages once I do use their junk for, say, a year [link] we might consider my claim that they have zero own skin in the game again, but, unfortunately, not before.

The type of failures I experienced:

• Electrodes intermittently did not work. Dry skin, whatever. They never really got there even with me making my skin humid, me rubbing my skin, or with me trying to soothsay the electrodes into submission.
• Limb position effect remains almost impossible to foresee and overcome. I extend the arm and the prosthesis blocks entirely, in whatever state it is, or, alternatively, it opens or closes against my explicit intention, whatever the state it was in.
• Gripper useless – iLimb finger configuration is just not good and failed over maybe 45% of daily grips. So I used an old Otto Bock hand and a greifer instead. The device still did not perform (see two bullet points before this).

Academic researchers and industrial developers surely must have seen that coming and did nothing to prevent that assessment / reality. There can be no surprises there. Au contraire³. I read papers where they seem to believe that these high  numbers for error rates can be “acceptable” or that we “negotiate” our lives in some way. They can now blame themselves, quite frankly. 

@article{castellini2020upper,
  title={Upper Limb Active Prosthetic Systems—Overview},
  author={Castellini, Claudio},
  journal={Wearable Robotics},
  pages={365--376},
  year={2020},
  publisher={Elsevier}
}

@article{chadwell2016reality,
  title={The reality of myoelectric prostheses: understanding what makes these devices difficult for some users to control},
  author={Chadwell, Alix and Kenney, Laurence and Thies, Sibylle and Galpin, Adam and Head, John},
  journal={Frontiers in Neurorobotics},
  volume={10},
  pages={7},
  year={2016},
  publisher={Frontiers}
}

@article{chadwell2021addressing,
  title={Addressing unpredictability may be the key to improving performance with current clinically prescribed myoelectric prostheses},
  author={Chadwell, A and Kenney, L and Thies, S and Head, J and Galpin, A and Baker, R},
  journal={Scientific Reports},
  volume={11},
  number={1},
  pages={3300},
  year={2021},
  publisher={Nature Publishing Group UK London}
}

@misc{gentile2023perspective,
  title={A Perspective on Prosthetic Hands Control: From the Brain to the Hand. Prosthesis 2023, 5, 1184--1205},
  author={Gentile, C and Gruppioni, E},
  year={2023}
}