Modifying an EEG headset for lucid dreaming

moddedmindwave4

[Michael], [Tom], and a few other people on the Lucid Scribe Database project have been using off-the-shelf EEG equipment to invoke lucid dreaming. Yes, that’s where you take control of your dreams and become a god. This requires wearing an EEG setup while you sleep, and these products aren’t very comfortable sleeping wear. [Tom] decided to take apart a NeuroSky MindWave and turn it into something comfortable to wear all night.

The folks at the Lucid Scribe Database log their dreams with consumer-level EEG equipment, usually something made by NeuroSky. The NeuroSky MindWave is the smallest and cheapest EEG headset available, but it’s still a hard plastic device not conducive to sleeping.

[Tom] removed all the guts and electronic goodies out of his MindWave and attached them to an elastic headband. The MindWave has two sensors – a forehead and ear lobe sensor. For the forehead sensor, [Tom] simply soldered a piece of wire to a penny and attached it to the elastic. The ear lobe sensor in the stock MindWave is a simple clip that was kept in the stock configuration for [Tom]‘s mod.

Now that [Tom] has a much more comfortable EEG setup, he can get on with improving his lucid dreaming skills and even try communicating via Morse from inside a dream.

26 thoughts on “Modifying an EEG headset for lucid dreaming

  1. Thats quite interesting, I would like to see interaction with the outside from within a dream but on the other hand kind of scary thinking about the implications..

  2. Very nice. I have been looking to make use of this headset for an Android app I wanted to write, and didn’t know it was that uncomforable. Thanks!

  3. Going to give this a try. I had 3 lucid dreams in the span of 10 days in my early 20′s and I would pay good money to experience it again. Don’t know what triggered them but so far the classic association trick failed for me.

  4. Hmmm, a little off topic, but have to say…. Never thought it was “lucid dreaming”, but I guess almost all of my dreams every night qualify as this. I almost always remember almost all of my dreams after I wake up, and probably every other night I have a dream where I try to read something. For some reason, I can never read anything in my dreams. From signs to a written letter I may be holding in my hand, no matter how hard I try, any type of cypher appears as blur. When this happens in my dreams, after several frustrating attempts to read and not understanding why only the text is a blur and not everything else, I think to myself inside the dream, “Oh yea! Duh, you idiot, you are dreaming, and you already know that you can’t read stuff when you’re dreaming, so quit trying and move on!”. And I do.

    My dreams are really exciting, but for some reason they are extremely rarely sexual or supernatural(flying, etc.), maybe only 5 or so in my lifetime. They are also not in real time, as it takes me about 5 minutes of real time to dream about one minute of dream “life”. Sometimes I can wake up in the middle of a dream to go to the bathroom, or get a drink, and then I can go back to sleep a few minutes later and pick right back up in the dream where I left off. Every once in a while, a dream from one night is a complete continuation from where the dream from the previous night left off. Even more rarely, I start dreaming a new dream than the night before, but after a few minutes into the dream, I remember(inside my dream, fwiwlol) that I was having a much cooler dream the night before, and decide to switch over to a continuation of the previous dream. I can also remember some of my most exciting dreams in great detail, some of them as far back as 20 years. And, finally, 9 times out of 10, I wake up at exactly the top of the hour, to the minute(I have the luxury of making my own hours, so I don’t use an alarm, and which hour it is can vary depending on when I went to bed, but still always at the top of the hour).

    I should volunteer for some sort of study, lol…

  5. Based on how this is wire up, this is more of an EMG than an EEG. Then again, it’s nothing new to see “consumer level” EEG devices that really aren’t reading brain waves.

    An EMG is of course better suited for the purposes of this hack anyway.

    1. What makes you think that it’s not an EEG machine? Did you study the design and so on? Oh man, some people really don’t get it. I know this comment is old, but I had to post a reply to this. I have a MindWave Mobile and it really reads brain waves. It really is not that hard to do this!

      1. Which wave is it monitoring? I’ve actually worked with real EEG and brain-computer interfaces that costs 10s of thousands of dollars, so perhaps it is you that doesn’t ‘get it.’

        While the mindwave may indeed have some super-secretly special sensor (which they claim to have and will release no details on), they position it in the wrong spot to extract a brainwave. Trying to access the brainwaves from the front have required superconductors and highly emi isolated rooms in the past. Devices that slap a skin conductive sensor on the forehead are measuring muscle impulses (Myo-electric potential) NOT brainwaves. Do you have any idea how weak brainwaves are even from the top? The signal-to-noise ratio is tremendous. Trying to pull a brainwave out of that is difficult even with equipment that costs 10s of thousands of dollars.

        The alpha mU wave which is the biggest and most easily accessible brainwave, for example, is mesaured from the top of the scalp, just left of center and represents the “will” to move a muscle.

        No, it actually IS hard to do this. I am guessing that you have never actually tried to build one. That’s usually what people say when they have never done it before. Build one and prove me wrong.

        1. I don’t have to prove anything. I don’t have time for this and it would be pointless. Single channel EEG can be built for $100 by an _amateur_. Just search for Olimex EEG, OpenEEG. Please Do it. Really. I was experimenting with my mindwave device and I was placing it in different spots on my head (yeah, it was ucomfortable). And I’m using eeg gel for this to be sure that a proper contact is there. And please… just by using a fact that something “costs 10s of thousands of dollars” you try to make me think that it can’t be built for a few houndred dollars? Sure, they may be using better electronics and so on. But a single EEG channel is just an amplifier + a piece of metal connected electrically to the scalp of your head with a conducting gel. Nothing more. Maybe there are some sophisticated machines with other electrode types but a traditional EEG device picks up an electrical potential from the top of your skin. Of course you will get lots of artifacts (search for “EEG Artifacts”) which you have to either filter out or be able to recognize on your monitor. I was sceptical about MindWave device as well but now I know it works. It picks up identical artifacts as a traditional EEG. It is showing my brainwaves going up and down (this is a simplified information) depending on what I actually do. And please don’t be deceived by the fact that medical grade hardware is so expensive. Its being tested extensively and it has a specialized software to make a life of EEG specialist a bit easier. Also it has to be expensive to make it “look” more valuable.

          And yeah I saw that ridiculous clip on youtube where some neurologist connected MindWave to a dummy which had a wet towel on its head. And he said “look, dummy head has brainwaves! hahahaaaa I got you NeuroSky!”. But be assured that medical grade EEG would have shown the same. Wet towel picks a potential from its surroundings. And since MindWave is capable of reading micro volts… it will pick up something! Even from thin air. I have an impression that you read too much and not actually experimenting with stuff. Because I’ve seen this argument where somebody was criticizing dry electrode from MindWave. But you know what? It can be dry because people almost never have hair on their forheads. If you want to place electrode somewhere else then you have to use gel. So this electrode itself _is nothing special_. You can replace it with a penny and it will work the same (besides search for EEG electrodes and you will see that these are a piece of metal with a gel chamber for better contact). The same with electrodes which are placed on the ear lobe. NeuroSky does not make electrodes secret. It’s the data processing that is hidden. Next to the amplifier there is that little ASIC module which has some analysis routines implemented. These routines recognize mind states and so on. They may be inacurate of course! They may be analyzing muscle contractions. Who knows. That’s their trading secret. But you can get a raw waves to analyze them yourself. I don’t understand how people may think that there is some kind of magic to it. Medical grade eeg = lots of “pennies” (on conductiong gel) conected to an amplifier + monitor to show the data. And if you really do your homework properly you will find other people experimenting with MindWave device with success and with the same conclusions that I have.

          1. Oh, and one more thing. You don’t have to contract your muscles on your forhead to make MindWave “think” that you are focusing on stuff. If I try to focus on something I already know then the “attention” metter in MindWave softwere does not go up. If I try to focus on something new which I was not trained with in the past … then the attention metter goes up. I’m not moving my forhead. I stay motionless all the time. Oh and it’s nothing new for traditional EEG to have electrodes on forehead (electrodes FP1 and FP2 used – for example – to detect epileptic seizures, but not only for this).
            Please refer to this drawing:

            You also can search for “EEG electrodes placement” and see where fp1 and fp2 go. Also find some EEG readings and you will see that these electrodes pick up the signal just as good as other electrodes. I think my reply is quite extensive and you see that this topic is not something that I’m unfamiliar with. I’m not sure what kind of EEG machine you were working with but maybe it was some old hardware. I don’t know. Please check this again because It’s hard to believe that modern hardware would have problems with this.

          2. You “don’t have time for this” yet, you posted a novel in response?

            Whatever. YOU have never built one, *I* actually have. So you have no idea. Done. Have fun.

          3. You do not know where electrodes are being placed in EEG (you said you can’t measure on forehead and I showed you that in traditional EEG it’s being done every day), you do not know how the electrode works. Why should I even believe you? You provided no facts which would cover your claims. I don’t have time for building stuff which has already been made. It’s just that simple. I bought a solution which was built by someone else. Man, if you can’t build it then don’t tell other people that they can’t. That’s the basic rule which moves this world forward. Besides this “novel” took way less time than building something and it was not even meant to convince you. It was to show other people that things are a bit different than you claim. If you’re building something – while being disconnected from the world at the same time – then it’s your choice. I just want to show other guys that MindWave works as an EEG hardware. And I don’t care whether you believe it or not. It’s your choice. I’m just sick of comments like “mindwave is a scam”, “it’s not an eeg”, “it generates random data”. And so on and on and ooon. And there are only few videos and sites showing that it really works if you know what you’re doing. You have built an EEG device? Cool, I’m happy for you. Unfortunately it is not sensitive enough to work on forehead (according to what you said). Fine. No problem for me. It’s just this one solution which failed. Move on, create something better now.

          4. If I may venture to get into this discussion, having used a wide range of devices and electrode types from BioPac to Emotiv to Muse to Neurosky, dry bare metal, dry Ag/AgCl pellets, dry conductive cloth with nylon and stainless steel, dry conductive cloth with nylon and silver,dry silver impregnated cloth, gel with Ag/AgCl, gel with gold, saline sponge with Ag/AgCl, and worked with older systems that used a grounded connection to the body to compare to modern systems which use a driven ground or Right Leg Drive to keep the common mode potential of the body as low as possible, and here are my conclusions.

            First of all, the Neurosky TGAM module (used in all of the neurosky devices, from MindWave to the Mindflex toy headsets to necomimi ‘cat ears’) uses a Right Leg Drive for the grounding lead, provides inputs for ground shielded electrode leads, input impedance of about 10^9 ohm, has an intrinsic noise floor of about 1 microvolt and a 12 bit ADC with 512 Hz sample rate and built in notch filter programmable to either 50 or 60 Hz, and outputs 8 normalized frequency bands, an ‘attention’ and ‘meditation’ value from a proprietary algorithm, and a ‘blink’ signal every second on average as well as raw EEG at 512 per second on average. The module operates on 2.8 to 4.2V with < 100mA current draw at full 56700 baud raw mode, 50mA at 9600 baud without the raw data.

            It is a truly excellent bit of hardware.

            That being said, electrode impedance and type of metal makes a huge difference. Silver/Silver Chloride has by far the lowest noise characteristics, and gel or saline produces a very low impedance connection. However, stainless steel dry electrodes are about as good as dry silver or dry gold plated electrodes and are not that far worse than dry Ag/AgCl. Dry Ag/AgCl are available fairly cheaply from BioPac and can even go through hair http://www.biopac.com/silver-silver-chloride-post-electrode. The high impedance of dry electrodes makes the signal unusably noisy unless a Right Leg Drive for grounding and ground shielded electrode leads are used. This is because of the huge common mode voltage induced from the surrounding environment.

            The MindWave electrodes are poor for a number of reasons, but not because of the material itself. First, the active and reference electrodes are too far apart for good artifact free recording from the frontal lobe. You need the reference at FPz, like in the Muse headset or the Zeo sleep monitor. Second, the RLD is placed on the other side of the reference on the ear. Since the RLD is actually driving the body at the average potential between the active and reference, it should be placed roughly equidistant from both, as they do with the Muse and the Zeo. My variation of this hack replaces the ear clip as well as the forward sensor, and uses three conductive cloth electrodes on the forehead, with the center tied to RLD and the others to reference and active leads using ground shielded cable. In this configuration, eye movement, blink and muscle artifact is greatly attenuated except for muscles directly underneath the electrodes.

            The signal quality is excellent in this case, equivalent to research grade recordings at the FP locations, and about as good as what the Muse provides.

          5. Marek, Justice099 is correct in saying that FP is the worst choice for a single channel EEG for sleep staging (it’s just fine for many many other applications however), but since that is really the only place that is easily accessible and with proper electrode placement and shielding it does work, that is why it is used. The higher you can place it on the forehead, closer to F3/F4, the better. You may want to experiment with the BioPac pellet electrodes placed on a band along the top of the head through the hair from C3 or C4 and Cz, with RLD again in the middle. I’m getting extremely good recordings and the comfort level is acceptable during sleep.

  6. so for the actual sensor that connects to your forehead you just used a penny? what material and design does neurosky use?

    1. The MindWave and Mindflex use a stainless steel dry electrode, essentially a bit of bare metal. The Mindflex Duel uses silver woven conductive cloth, very comfortable and quite excellent behavior for a dry electrode. The reference and DRL clip onto either side of the earlobe, and is stainless steel for all the above devices.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s