Biopotential sensors?

[BotB]

I've seen these used before for films like Beowulf where they were used to read and interpret the movements of the actors eyes for the CG (I do special effects so I kinda watch alot of extras)... Anyway I was wondering, do you believe it'd be possible to implement Biopotential sensors into a PICAXE system, or build your own Biopotential sensor for the PICAXE?

I'm nuts at electronics so I'd not know the first place to start looking to see if this were possible but if someone believes it is then I'll definately take up the challenge.

 

[westaust55]

Are you considering hooking the basic electrodes to a PICAXE - akin to using the PICAXE as an EEG/ECG monitor,
or are you considering sensors with some associated electronics and just want to read data from these?

Can you provide some links to datasheets for what sensors you are proposing to use with a PICAXE?

Which PICAXE chip are you intending to use?

 

[BeanieBots]

If the question is "can I connect a PICAXE to some bits of wire on the skin and get a reading?", then the answer is NO.
At best, all you'd pick up doing that is mains hum.

The more simple "is it possible?", of course, it's already been done!
However, finding a circuit or getting enough information to build one is something else.
If you browse around enough, you may well find enough information. If you find something but don't quite know how to implement it, post a link and we'll look into it for you.

 

[slurp]

I think most of these are implemented with "vision systems" tracking the eye balls, then relating the position to a direction. I guess the CMUcam might be able to do this but would expect it's likely to be a PC based solution with some other software.

regards,
Colin

 

[BotB]

So as a more universal answer I'd be looking for some piece of hardware that reads the "bio pulses" and outputs them as an analogue signal? surely then all you'd have to do is wire the analog output from the "said piece of hardware" to the analog input on say a PICAXE 28X-1 board? granted this is if this magical piece of hardware does output as an analog signal...

 

[BeanieBots]

So as a more universal answer I'd be looking for some piece of hardware that reads the "bio pulses" and outputs them as an analogue signal? surely then all you'd have to do is wire the analog output from the "said piece of hardware" to the analog input on say a PICAXE 28X-1 board? granted this is if this magical piece of hardware does output as an analog signal...

Once the "bio pulses" are EITHER an analogue voltage in the range 0v to 5v OR a digital signal of the same range, it's an absolute doddle.
The PROBLEM is getting from something as small and noisy as a signal from an electrode on the skin to something which can be worked on by a homemade circuit that does not make it worse!

If it was easy, then Nintendo and Microsoft would not bother with controls you need to touch or move. Just look at the screen to move the mouse!

 

[BotB]

Oh indeed I know it's not easy, I have been reading up on the subject. The most "reliable" way seems to be the use of EMG's, reading the pulse from a muscle. I've came across two useful pages.

This One gives a good overview of an EMG where as This one offers some insight into the EMG surface electrodes on offer...

Again I'm no pro but as you've said, it looks like the problem lies in amplifying the signal to 0v to 5v without upsetting the electrode... I also read that the 60Hz of most everything falls right in the middle of an EMG electrodes range...

 

[BeanieBots]

Interesting reading. To be honest, 100uV is quite a lot higher than I was guessing at. Removing common mode errors on such a signal should not present such a difficult task but removing "mains hum" without being able to use filter techniques is not something I'd like to try and do. It may be possible to use some form of "gaurding" technique but that would require individual setup each time and would probably need re-calibrating every time the subject moved. Maybe multiple "extra" probes to be used as artificial ground reference points??

I wish you all the best in your efforts.

 

[Ralpht]

100uV is reasonable but still very close to the noise floor. The 5mV that surface electrodes can detect is a lot better but still very low.

You will need instument amplifiers plus lots of analog electronics to reliabily boost the very low voltages to something that fall into the 5V range of an Picaxe. You will need to lift the signal from the noise floor, without carrying the noise and mains interference with it.

The 'big boys' do this with some very fancy (read very expensive) equipment plugged into fast PC's or custom designed computers.

Having said that, either Elektor magazine, Circuit Cellar or Silicon Chip, published an article on brainwave / muscle movement monitoring a few years back. The data was fed into a PC, not a controller though. Can't remember the article details, I have the mags somewhere at home - I think. That may be a good place to start at least.

I think though that you will need to collect a lot of data to get accurate and reliable info to track eye movement or whatever and the amount of data might be more than a simple microcontroller has space to handle.

Good luck.

 

[jglenn]

I did some work in the muscle stimulation field, worked with real docs and eng, making units that people with broken necks can use to operate their hands enough to grasp things. Wires were planted in the hand that caused certain muscles to move, the waveform was about 20Vac, a mA or less, short pulses of a few mS. Sensing nerve signals would be tricky.

One MAJOR point is that all medical eqpt of this nature is ISOLATED. There must be no path to ground or HV ac or dc sources. They are fanatical about this. People getting zapped does not help the insurance rates.

One idea might for small signal sensing would be to use a small battery powered module at the sensing site, then send the signal by radio or fiber optic cable to the processing unit. This provides for safety, eliminates ground loops and parasitic pickup. For eye tracking, how about a short focal length camera in some oversized eyeglasses?

 

[BotB]

Some interesting development has arisen through research. Turns out there is an OpenEEG "movement" if you will.

http://pceeg.sourceforge.net/
and
http://openeeg.sourceforge.net/doc/index.html

It seems people have figured out a board for reading EEGs, I'm interested to find out if this can be interfaced with the PICAXE. I have joined their mailing list and sent out a question related to this but I presume asking here and bringing this info to us all is worth the time too .

I read alot about isolating everything as you've said jglenn and production of this OpenEEG PCB seems decent, expecialy if you can get a useful reading from EEGs. It's not the EMG route I was hoping for but there seems to be more people looking into EEG so it seems like the best place to start.

 

[jglenn]

This will not be a simple or cheap field to experiment with electronics. Just picking up the signals is hard, they use some kind of silver discs with conductive paste on them. Then top grade instrumentation or opamp inputs. You will have tons of data and need to stream to a PC hard drive. Interpreting and processing it will be a chore. Please tell me this is a professional project and not hobby!

 

[Marcwolf]

Still...

It is an interesting idea. This was something I was considering when looking at controlling various circuitrs in a portable self-contain animatronics costume where the wearer needed to control the suits emotions without giving any indication that the control was happening to a watcher.

Such things like eyebroww raising and lowering to send a series of pulses, or the wrinkling of the forehead.

Or even the use of the thigh muscle to cause a tail to swing in syncronization with the persons walk..

I'll have to keep an eye on that forum.

Dave

 

[BotB]

I've had a few replies through the mailing list, one guy blew my mind but said that the PICAXE would be able to read the ADC from te Open EEG project but that it would be quire unreliable because of the build of the project board... It looks possible but not nearly 100% safe or reliable...

And on the subject of "animatronics" this was what i intended to use them for, if I could have this working before leaving university then I'm sure it'd go down a treat with employers.