How to Help a 99.9% paralyzed person

[Electrons-R-Fun]

Originally posted on the Savage///Circuits Forums
on January 27th, 2014, 04:32 AM
by ElectronsRfun

Hi I'm Jason,

I hope I am posting this in the correct location. This is all new to me. I thought I would just give a brief introduction into my life sustaining, life changing, and life prolonging project.

I have been working on what essentially started out as a call alarm for my wife who is 99.9% paralyzed but has progressed into the concept of giving her back a great deal of independence. I also want to engage her mind by creating video games that she can play and have some fun with. I was able to get her to play mind sweeper a little but she had a great deal of trouble moving the mouse cursors. Opps I got a little ahead of myself.

The real problem is my wife (Lisa) only has one discernible movement and it is about 1/4" of linear movement that's it. And it's not always reliable. So how do you harness this movement so a severely disabled person can do all the things I spoke about and regain a higher lever of independence back along with the will to live. I will discuss my approach to solving this problem the mistakes I've made, what I've learned ( mostly that I really enjoy electronics, I don't know anything , and components get really hot and leave burn marks on your finger tips when you hook them up wrong) hmm reminds me of the movie Indian Jones and the bar fight scene.) I have been doing electronics for about a year so I have a great deal to learn still.

Jason

[Granz]

Originally posted on the Savage///Circuits Forums on January 27th, 2014, 10:15 AM by Granz

Hey Jason,

Does your wife have control over her eye movement (left, right, up & down, as well as blink)? If so, Steve Ciarcia's Circuit Cellar magazine (IIRC) had an article on how to build an eye mouse. It allows the user to control the cursor movement, on a Windows PC, with their eyes. That might help. If that is something that you could use, let me know and I will try to hunt down more info about it.

Definitely sounds like a worthwhile project.

[Savage///Circuits]

Originally posted on the Savage///Circuits Forums on January 27th, 2014, 01:16 PM by Chris Savage with edits

Jason,

This forum is just fine for discussion of your plans and situation. If at some point you start documenting the project itself you can start a new thread in the Project Workbench forum here in the Bio-Medical Electronics forum. I believe when we talked on the phone we mentioned eye tracking and I agree with Granz that would really add to the capability.

[Electrons-R-Fun]

Originally posted on the Savage///Circuits Forums on January 27th, 2014, 04:32 PM by ElectronsRfun

Granz,

I would like that information definitely! I hate to bring up all the problems we have but my wife is down to one eye only now. Since your eye placement is controlled by muscles(ALS affects all nerve impulses to your muscles) her right eye has drifted out of center and causes her to suffer from vertigo. So I have to keep it covered. She also wears Rx glasses. Will the eye-gaze device still work under those circumstances? Also, is the device too large as where it obstruct her vision of the display she is trying to move the cursor on? The only thing she really has going for her is that she can watch TV. I could put the device on a servo or stepper motor so when she wants to use the device she activates the switch that turns the servo or motor on and rotates the device into place and when she done with it the device swings out of her field of view. We also have a 43" TV as the display so that should help with her twitchy eye-movements because the larger TV will "smooth out" her eye movement.

We have a Dynavox eye tracking machine. She can not use it very good. My wife's eye-movement is done more in jerks than a smooth glide like healthy people do. Her eyes often drift away causing her to look somewhere she does not intend and she blinks uncontrollably all the time and can not control her blinking. She also can not "stare" at any one point for a certain amount of milliseconds. I have tried all the variation on the Dynavox like: blink to confirm selection and stare at location to confirm selection, the only thing we could get to work for her was interfacing the LISA system into the Dynavox and she could confirm selection with the device I made for her. The interface includes a programmable blocking timing delay with visual red/green indicator to help user know when they are "staying on the switch" vs switch is ready to use again. I can post a picture or schematic later if interested.

Further Lisa can not use the Dynavox machine very long. This is typical of all people with ALS. Using the eye-gaze device is fatiguing and can only be used in short periods with long periods of rest. I would be curious to see if using a different style of device would have the same fatiguing effect.

Chris,

We did talk about eye tracking technology and if it was possible to use the prop-cam to do so. I wanted to track Lisa's eye and have a laser pointer connected to a servo point to the letter she was looking at to help her communicate with us faster. The letter would be as large and as close as possible to help with her eye flutter. My thinking on this would be from Trigonometry, larger angles from point of gaze would hopefully compensate for eye flutter by allowing a larger window for computer to calculate the general area someone was looking instead of the exact location of the gaze.

[Granz]

Originally posted on the Savage///Circuits Forums on January 27th, 2014, 11:32 PM by Granz with edits

Jason,

Take a look at the June 1995 issue of Circuit Cellar Ink (https://archive.org/download/circuit-ce ... ar-059.pdf) [Note: link is broken], page 20. I do not know if that would work for Lisa, but it would certainly be worth a look.

Another thing, to think about, would be one of those breath switches that some quadriplegics use. That would require that Lisa be able to control her breath and able to breath in and out to select different things on the computer screen. I have not seen anything about that myself, just heard about some devices (specifically wheelchairs) controlled by breath switches. I would not think that a switch, like that, would be too difficult to build, or to interface.

You said that Lisa has a tiny bit of control over one very small linear movement. What is the movement that she is able to make? Perhaps something like a Morse code keyer? The programming for that would require a tiny bit of artificial intelligence (take a look at the current keyer switch closed, or opened, time and compare it to the previous times to determine if it is a dot or a dash - you will have to adjust the allowed timing variance based on your wife's abilities, and the looser you need to make the timing, the less the reliability of Lisa entering the data/commands that she wants.)

[Electrons-R-Fun]

Originally posted on the Savage///Circuits Forums on January 28th, 2014, 12:33 AM by ElectronsRfun

Granz,

Thanks for the post. I mention A.I. in my patent application and definitely want to learn that. I think it will help if I just post my Patent application. I was thinking a combination of facial recognition and eyebrow movement would be best to simulate a person who is always watching over a patient like my wife. I envision that type of system to be almost human like and to be intuitive as to what the patient is trying to do.

Lisa is on a ventilator with a cuff, she can not use a puff switch. I will post my PPA on the work bench it is about 50 pages with the drawings but it will definitely show the direction I am heading and what I want to do. Lisa as well as most ALS patients can move their eyebrows a little that is what I harness from her to control everything I have gotten to do. I use a ratiometric Hall sensor to measure her movement. It is extremely accurate down to about 0.0013 volts with what I think you guys would call 12 bit resolution. I only recently learned to use the propeller a little after reading the first 118 pages of the PE kit. I don't want to write too much in this section so I will stop here.

I have to go pick up my son from college now I'll post later tonight.

Jason

[Electrons-R-Fun]

Originally posted on the Savage///Circuits Forums on January 28th, 2014, 09:06 PM by ElectronsRfun

Hi Granz,

I looked through and read the article you sent me. Reading about the discoveries others have made was fascinating. I think I can get this to work with my wife in some regards. I will build it soon and try connecting it to a propeller using the below style code to latch a relay. I was thinking PIN_0 could be part of the matrix grid switch Lisa is use to using to unlatch the relay.

IF ina(PIN) == -1
repeat until (ina == 0)
outa(PIN_0)~~

I have never built an amplifier circuit before, so I have never worked with negative voltages. I think they make negative voltage IC (like a 7805, but just in negative volts instead). I don't really know why neg. voltages are needed? To me it appears negative voltages only increase the Voltage potential difference. Anyone can jump in here and tell me why.

After looking at the schematic that was published I am not sure what everything is. I made a drawing of it and put the parts in red I don't understand. I hope it will be included with this post. I will try to explain what I think is going on.

First, I think the sensors are feed into a specialized comparator, the signal comes out and passes through a cap. The cap is placed in series so it would block all DC signals. I think this is called some type of band filter. Next the signal goes into a dual op-amp. In the first stage it appears to be amplified but I am not sure why R2 is connected where it is and then to common. I have never seen the symbol they use for common before either, is that because this system uses neg. voltages? Next the signal is feed into the second half of the dual op-amp where it appears to be amplified based on a voltage divider circuit connected to +. Finally the signal flows out of the op-amp and into something I have never seen before, R6 with a diode and cap in parallel on either side of the resistor. Huh... what is that???? Is the diode there because of the negative voltage system? The 3.3k resistor is part of a RC circuit that will charge C2 a little more slowly than if R6 wasn't there. I also looks like a low band pass filter, having the resistor before the cap. I don't know? As a whole the circuit looks very similar to a speaker driver circuit, never built one of those either.

I was hoping someone could explain the parts of the circuit I don't understand.


(Note: full size image lost during forum migration )

Thanks, Jason

Eye Activated Switches- Thumbnail.jpg

Eye Activated Switches - Thumbnail

(1.91 KiB) Not downloaded yet

[Savage///Circuits]

Originally posted on the Savage///Circuits Forums on January 28th, 2014, 11:55 PM by Chris Savage with edits

That circuits shows (in the first stage, including C1) an Active High-Pass Filter. The first op-amp is configured as a non-inverting amplifier with the gain set by R3 divided by R2 if I recall correctly (been awhile). R2 is not a pull-down and is part of a voltage divider made up of R2 and R3.

In terms of common, do you mean the ground symbol? There are a few different symbols typically used for ground. Technically each has a specific use but yes, ground potential is common and the supply is dual-polarity. The diode on the output of the second op-amp prevents the output from going negative.

On the subject of the ground symbol. The one used is Signal Ground. The one commonly used in its place is Earth Ground, though as I said, many use it instead. And then of course there is Chassis Ground. I have attached an image that shows each of them labeled properly.



I will try to follow-up again when I have more time.

Ground Symbols.gif

(1.86 KiB) Not downloaded yet

[Granz]

Originally posted on the Savage///Circuits Forums on January 29th, 2014, 12:42 AM by Granz

Originally Posted by ElectronsRfun
Hi Granz,
...
I have never built an amplifier circuit before, so I have never worked with negative voltages. I think they make negative voltage IC (like a 7805, but just in negative volts instead).
...

Jason,

Yes, there are negative voltage counterparts to the 78xx series of voltage regulator; they are the 79xx series.

Just like with the 78xx series, the last two digits will give the output voltage. For example, you would use a 7905 to regulate a negative 5 volts (-5VDC,) and a 7912 will output a -12VDC. Also, as with the positive-oriented 78xx series, there are sub-families; I.E. the 79L05 is the low-power regulator which will put out -5V @ 100mA. A normal 7905, as with the 7805, will allow you to draw 1 full A (different manufacturers variances apply - some will produce 1.5A and some 2A, etc.) You can see several examples of different regulators (including positive and negative parts) in Jameco's catalog at: http://www.jameco.com/Jameco/catalogs/c141/P13.pdf.

Originally Posted by ElectronsRfun
Hi Granz,
...
I don't really know why neg. voltages are needed? To me it appears negative voltages only increase the Voltage potential difference. Anyone can jump in here and tell me why.
...

As far as why negative voltages are required, those triangular devices are op amps (operational amplifiers) and those normally require both positive and negative voltages for operation.

[Granz]

Originally posted on the Savage///Circuits Forums on January 30th, 2014, 11:52 AM by Granz

Originally Posted by ElectronsRfun
...Lisa as well as most ALS patients can move their eyebrows a little that is what I harness from her to control everything I have gotten to do. I use a ratiometric Hall sensor to measure her movement. It is extremely accurate down to about 0.0013 volts with what I think you guys would call 12 bit resolution. I only recently learned to use the propeller a little after reading the first 118 pages of the PE kit. I don't want to write too much in this section so I will stop here.
...
Jason

Jason,

One more possibility: how about a camera with facial recognition? Hack-a-Day has an article (http://hackaday.com/2014/01/30/s-a-m-th ... n-monitor/) where a team from Chico State University uses OpenCV to track a persons eyes and give a driver a wake-up signal when the person appears to be dozing. You might be able to use similar techniques to monitor what Lisa does with her eyes/eyebrows/tongue/whatever and try to interpret what she is trying to do.