tDCS .. electric brain stimulation

2

Comments

  • Arduinos don't have an op-amp, so you'd still need the LM357/358.  Aside from that, sure, you could use an arduino, it's just a bit overkill.  Just make sure that you power it from a battery and that you unplug the USB cable, it is really important that there is absolutely no electrical connection between the tDCS and anything else.

    The op-amp based design is better than a potentiometer based one, not just in terms of cost, but also because the current regulation is handled by the circuitry itself and doesn't depend on the firmware to work right, so that fluctuations in impedance can be handled much better.

    I also have a great interest in ultrasound stimulation, though I have not had the time to do the research.
  • Why do you say Arduinos don't have an op-amp? Is it because the on board LM358 (on the UNO board) doesn't have dedicated in/out pins?
  • Oh, I wasn't familiar with the updated Arduino design.  You're right, the Uno does have an LM358, but one channel is already in use as voltage comparator (presumably to short-circuit the 3.3v regulator depending on the input voltage) and the other has its inputs tied to GND.  It's not intended (or possible) for you to use it for a custom purpose.

    The op-amps are cheap and widely available though, I can send you a few if necessary.
  • edited November 2012
    Thanks for clarifying that. The reason I'm interested in Arduino is not only is it pre-packaged for someone who's never mounted a chip to a board (like me), but the Mega boards are the perfect platform for my future projects combining EEG and tDCS in one device. 

    And thanks for your offer, but don't trouble yourself. I was just planning on buying some op-amps online. But heck, if you're giving them away, I'll take them off your hands! My current tDCS circuit employs a Radio Shack LM317 (which works great according to my ammeter), but obviously I'd like to create more sophisticated devices as I go
  • Remember that the arduino doesn't have any isolation features whatsoever.  Using it for tDCS or any biopotential measurement including EEG could be deadly if proper precautions aren't taken.

    I did once consider the idea of designing an arduino clone that does have full isolation and programmable gain amplifiers onboard for the sake of tinkering with things like this.  If there's enough interest, I could develop such a device.
  • I hear ya. I was interested in Openstim for a while, but the fact that the Arduino is connected to a computer during stimulation turned me off. If you're going through all that trouble, you might as well program a sketch that operates independently of the computer altogether.

    I don't want to make you busier than you already are, but I say such an Arduino project would be pure gold. With all the interest online these days, a legit tDCS design would be an instant hit with Arduino's availability, familiarity and resources. The whole reason I got on this Arduino fix is because I was looking at your circuit alongside an Arduino schematic, and they looked very similar. I would take on such a project myself if I weren't such a noob with no programming experience.

    I finally found a local store that sells LM334 & LM358, so I bought one of each to tinker with
  • I see that this conversation has been over for a few days. I would also be interested, however, in any results @rdb gets from testing.
    I have built my own (and used many times) tDCS device. It was an EXTREMELY simple model, as I am not an engineer (nor do I really understand electricity all that well). I received a lot of help from people on reddit. 
    I have also experimented a bit with brainwave entrainment. I have attempted on more than one occasion to utilize both forms of enhancement (tdcs and entrainment) at the same time, but with less than desirable results. I wanted to find a way to combine their effects so that I could design the ultimate enhancement helmet....but, I don't think the two technologies mesh.
    Anyway, there's my two cents. :)
    (P.S. I still engage in stimulation with tDCS on a fairly regular basis. I'd have to say it is my preferred method of biohacking and research into the subject was how I found this forum).
  • RDB,

    I too would be interested in buying a kit or two from you as well.

    Also, if it's not too much trouble could you keep me informed on your progress?

    [email protected]

    Thanks much!

    Robert

  • @wrobert i guess updates will be posted in this thread. if you like to receive email notifications you can "favorite" this discussion and enable email notifications about new posts in favorite discussions in your account settings.
  • edited November 2012
    Anthonylee--

    I'm also quite interested in using tDCS to enhance the effects of brainwave entrainment. In my earlier posts I talked about oscillating tDCS for this purpose. It looks promising, I recommend you look into it. You may also be interested in Neurofeedback (Peniston-Kulkosky protocol), which trains you to enter different brainwave states at will without using external stimuli

    Also, I ran across a study that combined tDCS with TENS showing that the two methods together have a greater effect than tDCS alone. While TENS and brainwave entrainment are pretty different, I think the general principle is that anodal tDCS makes neurons fire more easily and therefore ANYTHING your stimulated brain is engaged in will be enhanced. So far I haven't found any research on tDCS alongside brainwave entrainment, but my guess is the two methods will complement each other. Also consider tDCS is localized to specific brain regions while entrainment is more of a whole-brain, interhemispheric approach. My semi-educated guess is placing the anode over the auditory cortex will enhance the effects of binaural/isochronic tones while an anode over the visual cortex will enhance light pulse effects. 

    Here's the study:

  • edited November 2012
    Here's a must-see video of an expert from Soterix Medical giving a tDCS lecture (at Harvard I think). It talks about HD-tDCS electrode placement and its advantages over the standard montages.

    https://clinicalresearchlearning.adobeconnect.com/_a1029693963/p9c2yu065w6/?launcher=false&fcsContent=true&pbMode=normal
  • I stumbled across this video this morning or a rather advance Italian tDCS machine.  Apparently they are not fucking around.




  • @rollsroyce Excellent. All this time I have been focusing my neuromodulation methods on the results I was expecting, and not paying attention to synergistic electrode placement. The auditory cortex. The answer was right in front of me the whole time.
    Also....I would LOVE to get my hands on some HD electrodes. Or an HD tDCS device at all.

    Thanks for sharing. :)
  • As far as I know, Ag/AgCl disc electrodes as used in EEG/EKG are optimal for HD-tDCS, and they are widely available online.
  • I studied the following video to get a better idea of how to place leads according to the 10-20 eeg method.  The accent is heavy, but the notes on the screen make up for it.


    and combining it with this excellent webpage that most of you have probably read.  If you have not read it, I found that I needed to pay close attention to figures, 4, 5, 6, and table 2.


    Here are my notes from the video to save you some time, but you will still have to watch most of the video.

    Nasion, is the point between the forehead and the nose.

    Inion, the lowest point of the skull from the back of the head and is normally indicated by a prominent bump.

    Left and right preauricular, near the center of the ear length, in front of the small skin ridge affront of the ear.

    Cz (Central Zero or central zone):  the distance from nasion to inion is measured in centimeters and is the total divided by two. 

    Fpz: 10 percent of the total distance from the Nasion to the Inion measured from the Nasion toward the Cp

    Oz: 10 percent of the total distance from N-I from the Inion towards the Cz

    Each electrode site has a letter to identify the lobe along with a number to or another letter to identify the hemispheric location.

    Even numbers identify electrode positions on the right side of the head, and odd numbers refer to the left side.

    The label "z" points to the three electrode sites in the midline of the head:  Fz, Cz, Pz

    F stands for frontal
    Fp stands for frontopolar
    T stands for temporal
    C stands for central
    P stands for parietal
    O stands for occipital
    A auricular (ear electrode

    T3 is measured by using 10 percent of A1-A2
    C3 is measured by using 20 percent of A1-A2
    T4 is measured by using 10 percent of A1-A2
    C4 is measured by using 20 percent of A1-A2

    The longitudal Circumference is measured along Oz+T3+Fpz+T4
    Distance from the Frontalpolar point (Fpz) for Fp1 and Fp2 is 5 percent.
    Distance from Oz for O1 an O2 is 5 percent.

    F3 and F4 are located by point of intersection.  

    P3 and P4 are located by point of intersection.

    Electrodes placed on ear lobes or mastoid are called A1 for the left side and A2 for the right side.  


  • @rdb, but every time I see pictures or video of HD tDCS sessions, it appears they have several small disc electrodes placed around the stimulation area. I don't know how I would get my device to evenly distribute the current through many electrodes (much less even attach them all).

    @jlwcamus, from what I have learned from the guys at brmlabs (where I learned almost everything I know about tDCS) unless you are using HD tDCS (which uses smaller electrodes on very precise points) the stimulation area doesn't necessarily have to be so precise. This is because most sessions use sponge electrodes around 2x2 to 3x3, and they disperse the current over a pretty wide area anyway.
  • rdbrdb
    edited November 2012
    I'm not sure about that either, but I imagine that just tying together all of the anodes (or cathodes) should work.
  • @rdb

    Hi I was wondering if you could possibly send me a copy of the schematic tdcs4 ? When I click to enlarge it here I can not read any of the specifics the pic is too small. This is in reference to your may 28th comment.
    Thanks much! My email at home is [email protected]

  • I also think you can pull of HD tDCS with a single channel device by simply having one wire connected to all four of the outer electrodes (my guess is the pros use multi-channel devices for an added measure of safety/control). If that doesn't work & the majority of current runs through only one of the four electrodes, the current can be evenly divided using resistors. If that still doesn't work, I guess multi-channel devices are the only option

    rdb, do you have to put sponges under the AgCl electrodes, or do they work by themselves without burning your skin? 
  • @wrobert: You can just use the direct link: http://rdb.name/tdcs5.png

    @rollsroyce: Right, that's what I meant.  I believe the AgCl electrodes are used with electrode gel, but I'm not certain.  Check out this paper:
  • So....taking on the task of building your own device is pretty awesome, and definitely something a biohacker would find fun and challenging, however for anyone who has the expendable cash I wanted to share this:

    It is a pretty nice unit made in China. I tried doing research on the company but it seemed too shady for me to risk almost $400. However, I am an active redditor in the subreddit for tDCS and someone has bought one and claims it is awesome. 
    Here is his review 
  • I was just at a Quantified Self meetup in Boston. I mentioned that I was considering experimenting with tDCS to a group of people there.

    It turns out that the guy standing to my left was an instructor in neurology at the Berenson-Allen Center for Noninvasive Brain Stimulation at Harvard, where they are experimenting with none other than tDCS.

    So I directed him to this site. He seemed excited to hear that this was going on.

    Networking, ftw.
  • @rdb, I read something you wrote on this forum about a design you had for a tdcs device and the possibility of manufacturing it. I couldn't leave a comment at the time but I am very interested. thanks.
  • I'm sorry, but I've been ill during the past few months and it's not likely that I will recover until a few more months, so I can make no promises whatsoever.
  • edited March 2013
    Hi 
    I just joined to make a comment.
    I have been designing medical instrumentation for 15 years and have a masters and PhD in biomedical/electronic engineering. I specialised in ECG circuit design.
    Firstly, sorry to hear you are not well rdb.

    But now, to the point...
    There is little wrong with the initial circuit provided by GoFlow.
    The LM334Z will not be damaged by removing the battery (which incidentally is equivalent to turning the power switch off). Doing so makes the return path open circuit so no current flows and nothing gets damaged. Even if the battery is accidentally shorted the LM334Z is specified to withstand 20V of reverse voltage if you check the data sheet, so 12V reverse voltage from the cap is fine. I would only change the design slightly - probably by swapping the fuse for a 1K resistor, adding a film capacitor across the electrolytic and putting a 100K bleed resistor across the capacitor. The main function of the 1K resistor is to act as a current limiter both in forward and reverse direction. Now there are maybe better ways to limit current but a resistor is fine and is easy and cheap. A 1K resistor will limit DC current to ~10mA, which although is a bit high for extended use is not life-threatening. The way the design is, with the fuse, if you turn the power on before putting it on your head the capacitor will charge up to 12V. Then if you put it on your head and the electrode resistances are abnormally low, you could easily blow the fuse (even if just by accidentally shorting the two electrodes). This would be a PITA. 

    Now rdb's alternative circuit is not as good really. It is more complicated than it needs to be, needs programming and is not protected from electrostatic damage. Any high voltage across the inputs is shorted directly to 0V by the varistors, causing high current through the electrodes, which could hurt somebody if they were touching one of the electrodes. This voltage will also appear at pin 2 of the opamp, possibly blowing it up as well unless the varistors have low clamping voltage.  This can be fixed by putting series resistance (about 500 ohms) in series with the electrodes. This basically limits the input current to the electronics and through the electrodes. The opamp will itself clamp its input voltage to its supply rail via internal protection diodes (so functions as the varistors are intended to) - but you MUST have series blocking resistance to restrict this current. This is standard practice in any circuit connecting to something else by a connector but essential for something connecting to a person.
    Just throw away the varistors - they are not needed unless you want to make it defibrillator proof. Even then they would normally be used in a T configuration with blocking resistors on each side.
    Sorry to be a pain but IMHO the suggested improvements were a backwards step really. Anyone intending to build this is better advised to go for the simpler circuit of GoFlow (maybe with changes I suggested).
  • rdbrdb
    edited March 2013
    No, you raise some good points, and it's good to be open about apparent shortcomings so that the best design can be reached.  (To think I had almost rejected your application for being too generic...)  I also concur with your suggested improvements to the GoFlow design, although some of my concerns with it still stand.

    About my design, there is actually a resistor in series, although of relatively low value.  It serves to reverse bias a FET, which blocks when the voltage drop across the resistor becomes too high as a result of a too large current.  Having played with EEG/ECG designs myself, I'm aware of the practice of placing resistors in series with the inputs. In the case of EEG/ECG, the input impedance of the op-amps is so high that a few extra kΩ doesn't hurt.  However, in the case of tDCS, adding a resistor large enough to sufficiently limit the current will significantly increase the output impedance and therefore require the use of a higher supply voltage in order to be able to achieve the currents required for tDCS.
    Could you explain why the FET-resistor setup is not sufficient to limit the current?  Could it perhaps be adapted to function appropriately without causing a significant increase in output impedance?

    The primary goal of my design is for the signal to be controllable by a microcontroller, so that the ramping can be tightly controlled and that the session can be programmed (and the exposure time limited).  I consider this an important safety feature as well, as it protects against user error (turning up the current too quickly or leaving it running for too long).  It could also serve to shut down the device in case it detects sudden drops in output impedance or something of the sort.
    Not only that, but the device could also easily be adapted to produce more complex output signals, such as one required for oscillating-tDCS.

    You raise a good point about the varistors.  A friend of mine who's an electronics expert had recommended putting them there to protect the circuit against ESD caused by the body being charged, which might result in the circuit being damaged and a too large current to flow.  I'll look into T configurations.

    Anyway, the criticism is greatly appreciated, and I look forward to reading your future suggestions and comments.  I will not release a device as long as there is unresolved criticism.
  • Hi, I have just joined and I find the discussions fruitfull. I have followed the discussions above and need to ask the question, how does one ramp up the current and ramp down in the circuitry. and to include a timer into the goflow design.
  • I'm bumping this topic because this now exists. According to the company it will start shipping in July--and it's reasonably priced.
  • $250 is not a reasonable price for a tDCS device.  I'm guessing it's so expensive because of all the non-essential extras (bluetooth communication, iPhone app, etc.)
  • Yeah.  Even bluetooth support, etc. isn't really a justification for that price; that's still a huge markup.  If they also had things like other forms of stimulation, EEG support, etc., that price might approach something like being reasonable (and it would also justify having an app in the first place).
Sign In or Register to comment.