Check out this customer eye-tracker project!
Well, we are back from the holiday and are ready to start the homestretch until Christmastime. We hope you all had a great Thanksgiving! Today, we're talking about a great project from SparkFun customer Luis Cruz - the Eyeboard.
Based on the ATMega328, the Eyeboard is an open-source project built with the intention to help people with disabilities. The board allows users to control a computer interface using only their eyes. While projects like this have been made before, this project is made more impressive because of its low-cost, simplicity, and - well - the fact that Luis is only 18 years old. The system uses electrooculography, also known as EOG. EOG essentially uses electrodes placed around the eyes to record eye movement.
Check out Luis' website for more information or to order your own DIY Eyeboard kit. As the Eyeboard is open-source, Luis has links to download schematics and code as well. Great work!
LOL your username is noworries and you're worried. Now that's funny.
Why isn't this up higher?
I keep stressing this to hobbyists, and it's safety!
I think the probes are a bit like capacitive sensors, and don't pass any sort of current that's hazardous. Licking a 9v battery is going to be a hundred times more dangerous than this, and that's kinda safe. (PS: those that haven't licked one, try it out some time.)
I think the probes are a bit like capacitive sensors, and don’t pass any sort of current that’s hazardous.
Whoa there! Capacitive sensors may be capable of passing an AC current that could be dangerous. DC isn't as bio-hazardous as AC can be.
There are two types of probes, and, without donating $100 to the Kickstarter project, I can't determine which it is (which is a terrible reward system, IMO). I'm going to guess, though, that it's not the capacitive type. For most disposable electrodes like those shown in the 5th image, this is a silver-plated electrode in silver-chloride gel. Also, capacitive signals are much lower-level (more than $125 will sense), and are not very useful on low-speed signals like EOGs, which can be <1 Hz.
When you bypass the protective layer of the epidermis with conductive gel, your body's impedance drops tremendously and a few hundred millivolts can be increadibly painful. I use current-limited stimulator bars to stimulate muscles, and 25 mA is about the maximum I can take before it becomes really painful. A 9V battery across these electrodes could cause out-of-control spasms.
you could get muscle tetanus ( sustained contraction ) with a nine volt but not spasms, this would need to be done using alternating current.
I don't really know how these work, but all I was saying is that I thought they were passive sensors, that didn't get involved electrically with the 'sensee'
"Licking a 9v battery - try it out some time."<br /> Tastes like burning!
I've licked a signal generator at 16v, and tried different kinds of waves! </br> Squares and sines taste very different to me :P
Ok, I'm gonna have to try that.
I hooked a variable power supply to my tongue piercing and hit it with 12 v, I felt the direction of the electron flow, reversed the polarity and felt it go the other way...totally cool, till my classmate reached for the dial and spun it...I yanked the probes off the piercing in time, hehehe. Good times in college...
I think it might be cool to lick some music! I could just hook up 2 probes to a headphone output, and taste some I Follow Rivers by Lykke Li! And some of the music I've made, and some Astral Projection... aw, heck, how 'bout my entire music library!
With out seeing the schematic it is hard to know what is going on in the system but most op amps will have internal diodes that would prevent back current, also if it is powered by a battery it is normally considered safe, at least according to the instruction that I have had (BS in Biomedical engineering May 2011) . if connecting to mains power it would be important to use electrical isolation and appropriate ground pathing. this can be done in several ways but normally done with opto coupling between the adc and the MCU if using an external adc, or on the power lines if using a on board adc.
I contacted the author, he says that schematic is not available just yet, he only have a DIY kit that will be selling in the next year, they are going to be available the by then.
You can also check out his Kickstarter page, where's he's trying to get the funds necessary to manufacture it. You have until December 22 @ 4:15PM EST.
Thanks for your comments. For those of you that are concerned about the safety issues about the circuit... Well first of all keep in mind that the high impedance of the Op Amps makes it almost impossible to electrocute somebody.
We have tested the project with many users, and even disabled people have already used this system. We have made sure that it is safe to use it, and when I built The Eyeboard, my main concern was to make the circuit totally safe for the person that uses it.
CNN covered the story as well, and you can see someone with disabilities writing with the eyeboard: http://www.youtube.com/watch?v=r8d6j8b3sGI
Also, an EOG, unlike an ECG, is not connected to a fragile part of the body such as the heart so it is totally safe to wear the electrodes. Only the ground electrode is connected to a person's forehead, but it is NOT a requirement, it can be connected to the user's hand for instance.
MAKE Magazine is going to publish a "How-To" article next year, they replicated the project and they know how safe it is to use The Eyeboard, they are probably going to carry some kits to sell them on Maker Shed!
For more information go to my website: http://www.intelsath.com, and if you are interested in supporting this project that can surely help a lot of people, you can contribute by donating on my website or at kickstarter: http://www.kickstarter.com/projects/lcruz/the-eyeboard
INTELSATH Chief Executive Officer http://www.intelsath.com MyICV Resume: http://www.myicv.com/lcruz
Want to pre-order an Eyeboard Kit? Shipping starts on February 2012! Pre-Order now at http://www.intelsath.com
Support the Eyeboard at kickstarter: http://www.kickstarter.com/projects/lcruz/the-eyeboard
Remember to tell your friends about this unique kit that can surely help a lot of people! See video: http://www.youtube.com/watch?v=zObidmxnj5Y
IEC 60601-1 covers the safety requirements of medical electronics. You should read it. The issue with patient connected electronics has to do with isolation from mains power. Most electronic devices that are powered from the AC line are isolated from the AC line by a transformer. If that isolation fails, the connected patient can be electrocuted. IEC 60601-1 requires an additional isolation barrier for patient connected electronics.
Powering the device from a battery and not allowing any means to charge the battery while connected to the patient would get you around the isolation issue.
You may want to contact the manufacturer of the operational amplifiers you use and ask them if they are willing to guarantee that their products insure safety as you think they do.
You may also want to review the datasheets for all the components you use to check for restrictions on their use in safety critical applications.
Keep in mind that operational amplifiers can and do fail, and one failure mode is a dramatic drop in input impedance.
Also, when you state that you connect a ground electrode to a person's forehead or hand and feel that this is safe, I can't disagree more.
Unless this is done through a high impedance, dangerous currents can flow through the human body if they come into contact with an external voltage source.
The fact that you have used the circuit on humans and nothing has gone wrong yet does not prove that the circuit is safe to use.
I would suggest you make a complete schematic available to those interested for review, or that you contact an expert in medical electronics design to assist you in making your product as safe as possible.
I applaud your desire to help those with disabilities to better interface with the world, I wish you the best of luck and hope you do so safely.
If you put electrodes on a body, you should assume they are placed across the heart. -Or someone will put them there. You must protect against all foreseeable modes of misuse, and failures, all happening at once.
I'm sure your design works fine, and is safe under normal conditions, but the safety issue is about what happens in a fault condition.
I don't mean to be discouraging, but I would be advise caution before anything for sale. Ethics aside, this might take you into a position of legal responsibility. Maybe that's why you dont see many kits of this nature.
And I'd also second noworries's comment about applauding you doing this type of project. Industry tends to neglect these low profit areas, and seemingly small things here can be life changing for people - unlike the latest remote controlled touch screen widget fad.
I really recommend reading the iso standards for medical equipment. They are incredibly long, dry and boring - but every rule is there for a reason, and I certainly learnt a few things from them. Sometimes they seem prohibitive, but its more about producing a carefully considered design rather than something with multiple cages around it.
+1 for using Linux! cool and potentially very helpful project.
Linux is /dev/cool!
...And those of us that use VMs know that Windows belongs in /bin.
[if cool is in /dev, can you mount it??? :-P]
Like the garbage bin! Hahaha!
That made me think of rainbow dash.
(I don't use a VM, but I'm pretty sure Windows doesn't go in /bin - it's not an executable and it certainly isn't required for you system to run.)
Was wondering how long it would take someone to pick up on that. It's just the name of the folder -/bin - that makes it funny. Haven't fished around for it, but it probably goes in /srv or /etc
On his website he says that this is completely open source but i see no way to get any schematic and pcb layout without paying. Im sure he used a lot of info from other open source projects so the least you can do is share it with the community easily.
Easton, I am not asking anyone to pay to get the schematics, I am only asking for donations and pre-orders for anyone interested. You are not getting any schematics right now even if you "pay", as I don't make them available just because I am working on them an and a DIY guide to make it totally available on my website when they are ready. However, I will publish them when the kit is up and running. :)
im confused what this is then on your kickstarter:
Besides being a featured sponsor, this gives you access to the premium section of our website, which lets you download schematics, and code of other projects other than The Eyeboard not available for other users! One of our most exclusive projects, is a microcontroller video game system that features a wireless motion controller (The EES).
if those are already free then what are you paying for?
Let me clarify. I do have lots of more interesting projects, which I do have them on my website as videos. One of those is a video game system I developed (EES), but I haven't released those as open source yet... So people that donate at kickstarter, one of the rewards is the access to those other schematics and code (which are NOT open source yet). Like I said in the description, "...which lets you download schematics, and code of OTHER projects other than The Eyeboard".. I am sorry if the description is not so clear, I will make sure to revise them and add a better description. Thanks for pointing that out though! ;)
Without a schematic and pcb layout, it's anyone's guess about the safety issue. Such a device MUST be isolated, even for the diy market. Would be pretty easy for diy if it were battery powered - just add optoisolators for comms and keep the two sections well separated on the pcb.
Luis, if you read this I think you'd be wise to read ISO 60601 before offering anything for sale.
Safety issues aside, gotta say I think this is an awesome project.
Isolation should occur no later than after the preamp and possibly the lowpass filter. The energy contained in the batteries would still be dangerous. It's good of you to encourage caution, but don't advocate solutions unless you know what you're doign.
The purpose of isolation is to float the electrode wires with respect to mains ground.
As such, it not so important where the isolation occurs, so long as the patient side part of the circuit is isolated. Many commercial biopotential measuring systems isolate immediately after an ADC. Analogue amps ('modular' research lab type gear) tend to isolate as soon as a low enough impedance signal is available, usually very close to the input as you say. It's about what makes sense in the design.
Granted, electrical faults within the device itself must be considered, but the bottom line is when electrodes are connected to a subject, you must assume any and every component can have a catastrophic failure, and protect against all combinations of 2 (I think..) concurrent failures which could deliver unintentional current to the patient. - so having a single stage preamp or a multi stage amp with ADC present a similar hazard. Although isolating earlier does reduce the chance of 2 concurrent failures, it may not change the protection scheme.
Agreed on the impedance(not resistance) of electrodes going quite low, and some faradaic behaviour occurs in any electrode, but a 9V DC battery directly across these electrodes –could it kill? Assuming 50 ohm resistance per electrode, 2 electrodes, 9v/100 = … it might hurt, but..
Having said all that you do have a point, these designs must be well thought out, and it safety goes beyond just isolation.
NEC requires that ground fault interrupters trip between 4 - 6 mA, so even the 90 mA that you might expect to be possible assuming 50 ohms resistance per electrode is greater than that by a factor of 15.
Not safe in my analysis.
There are ways to avoid low impedance connections to the front end of this circuit which will enhance the overall safety of the device.
For instance, one can have large series resistors that feed high impedance instrumentation amplifiers that provide significant safety enhancement for the user in the event of a circuit fault. Since these resistors have almost no current flowing through them in normal circumstances, they also will have almost no voltage drop across them.
This is not to say that just to add these resistors will make the design safe. A total analysis of the entire circuit, including any connections to exteral power sources and computers is required, along with the printed circuit board layout, and the project enclosure.
Failure to do so may lead to bodily harm, and can also lead to liability lawsuits that may have a negative impact on one's career.
This could have tremendous military applications. Good work dude
I looked for a complete schematic for his circuit and did not find one.
I would be very hesitant to connect probes to my forehead that are interconnected to a laptop without a detailed analysis of all the possible failure modes of the circuitry involved.
I.E. If an ESD event occurs, is it possible for the op-amp's input impedance to dramatically change allowing a path for dangerous currents to flow to one's head?
Please note that since I have not seen the complete circuit he is using I can't say one way or the other that it is safe or not. I am just saying that I would want to be very sure that I did carefully look at the circuit before hooking myself to it.