Author Options:

I want to use nerve impulses from muscles as a data input. Answered

I need to find out how to get a nerve impulse isolated and use it as an input.  I beleive a myoelectric electrode with a window comparator may work.  Instead of using an op amp to boost the impluse to run a motor, I want to use the impulse as a data entry signal. 
I know that prosthesis use this system and I want to adapt it for signal input.  My goal is to use several nerve impulse sources to be used as a data input device to allow those with CP, or amputations to more easily communicate.  I am trying to bring a very old idea I had 20+ years ago to reality.  I have joined an open source prothesis forum as well as an Arduino forum.  I want to use the Arduino to "translate" the input combinations into a data form that a computer, PDA, or smart phone could use.  I want to keep this open source as I am not interested in a commercial venture.

There are basically two steps. 

The first is to use an Arduino to convert 5 touch pad inputs into 120 "characters" that are identified with various platforms.  5 inputs will generate 120 combinations (5x4x3x2x1) and that will give me a simple keyboard input.  I have to work up a timing method so that the Arduino will identify when buttons are pushed individually vs at the same time.  Next to have the Arduino generate a digital output that is compatable with a keyboard to keep the application flexible.

The next step will be to identify intended muscle nerve signals from background nerve signals and 'filter" out unintentional signals.  Next, the buttons with the myoelectric inputs as the input to the Arduino.

My goal is to have a myoelectric (or similar) input device that does not need muscle movement to generate data that is recognized by (practically) all devices.  It could be used for text generation, then voice synthesis, as well as a data signal for motors on the prosthesis to give better control.  Touch screen devices are proving to be a big challenge for those with prosthesis.

Thank you,

Paul H



Best Answer 7 years ago

There are a couple of Instructables on this subject, if I recall correctly. Look at the I'bles in the Assistive Technology Group.

The introduction to that Group will also point you to both user accounts here, and organizations on the Web, who should be more able to provide you with assistance. In particular, look at the Open Prosthetics project.

I tried a simple Google search for "DIY myoelectric pickup", and the very first hit has lots of useful information, including a schematic.

Part of the "art" of Web searching is being flexible with your search terms. If you don't find what you want the first time, try reading some of the results and substituting terms from those if they seem like what you want. Or add additional terms (like "DIY") that narrow the focus.

Thank you for your suggestions,

I have joined the Open Prosthetics project and I am still searching. I have also joined the Arduino forum as well as a volunteer program that matches up engineers with handicapped paients.
I think that I am going to have to be a little more patient.

Just picked up an Aruino UNO tonight and I am plugging away at the basics so that I can sort the logic end out. This Arduino is a lot of fun.

I will try "DIY", I forgot all about the " " in the search.

Thank you all again, I'll make a coffee and slow down a bit.

Paul H

All I can say is wow and good luck to you :-).

I do have progress on the logic side as I can use the Arduino sketch for GKOS and modify it for my application. I will have program for a bounce though. Baring the usual "ooops" that happens with a noob programing something I am 100% sure that the logic can be done.

My biggest challenge, which I thought would be the easiest due to the fact that myoelectric signal filtering is not new is to get a myorlectric filter device of schematic.

Penn State just developed one for input from a person's temple so I know that what I need is currently being used.

Yes, I will run the filters on batteries and will probably use a blue tooth Arduino just as if you would a blue tooth wireless keyboard.

Thank you for all of your inputs. I am not going to let this 25 year old project end up on the shelf again. There are too many people that would benefit and the cost would be a tiny fraction of what is currently available.

Take care,

Paul H

great I think it will definitely help people but why don't you run the filters on some other energy source besides batteries so that way it is not only assistive but "green" too.

Making this green is a great idea but I have to solve the fundamental problems first.
Solar panels to charge batteries are probably the easiest route to go as it has been done several times.

Thank you,

Paul H

I understand your comment.

I have been searching for quite some time but I end up with medical papers that are way over my head.

I have joined volunteer associations as well as open source forums but so far I have had no luck in getting my hands on a schematic for the input. I know that there are several devices out there as this is not new technology.

I will continue to refine my searches but I was hoping that I would find someone to point me in the right direction here.


Actually, what you want here is EMG -- electromyogram.

Usually what's done is to measure a set of muscles which are still under voluntary control but aren't normally used for much. The most recent version I've heard about used some of the chest muscles, if I remember correctly. Then count on the fact that humans are pretty darned good at learning to associate actions with results.

I'm just thinking of circuit techniques - any half decent instrumentation amp will do the job really. I was looking at ECG circuits not long ago, I haven't looked for EMG, but I can't see the techniques being that different.

And of course we MUST talk about circuit isolation.

Isolation: Absolutely. Safety first.

Amplification: The key point, if I remember correctly, is that the signal levels are low enough that you need pretty hefty filters to keep 'em from being overwhelmed by ambient noise.

Nite Owl: "The first time I used it, it broke my arm. Never again."
Silk Spectre: "That sounds like the kind of costume that could really mess you up."
Nite Owl: "...Is there any other kind? "

As I see it, the myoelectric signal isn't much different to an ECG signal, the key is multi-electrode techniques, so you only extract the signal around the muscle, not the whole body.

If you websearch on assistive technology, prosthetics, and similar keywords, you'll see that there has been a lot of research in this area. Since much of it has been documented in published papers, I'd suggest doing some literature research rather than attempting to create this de novo.