MEA Project - call for collaborators

2

Comments

  • The one that Mark Gasson used, I believe, only copies itself onto any network that scans the tag and then downloads itself into any tag that's scanned by that network.  I can't imagine you'd be able to do much more with a virus that can fit into a typical NFC tag.
  • IanIan
    edited May 2014
    So, Gasson hasn't responded.  I just sent him another email in case he didn't see the first one, but I'm not really optimistic at this point.  It shouldn't necessarily be that much of a drag, though, as I can get a decent idea from looking at some of the papers what some of the design actually entailed.  To an extent, though, we still will have to end up reinventing the wheel if we don't get a response.
  • Also, I just noticed that the OP asked if anyone would be willing to be the first test subject.  Put my name down for that.
  • Bastich! Didn't I already volunteer? Fine, I'll wait for the cancer-free version.

  • Who says we can't both do it?  Arguably that would be better anyway.  Though I assume that we wouldn't both get it for free.
  • Well, its still in the planning phase. I think it depends on which components are made and which are purchased. If the Array alone is 4k, then two are out of the question unless we get significant further funding.
  • @Ian Mark Gasson works at Reading university, we could probably track down a telephone number for him... if we can find someone near enough we could even visit him.

    As for the cost, yeah, the current funding will probably only cover one device. But if we are successful, I am willing to provide more money to improve it, plus with one success to our names, will probably be able to get funding from other sources.
  • I pretty much have zero expectation of any response from Gasson at this point.  If anyone does contact him about this, it would have to be by phone or in person.
  • someone get me a number. i will do this thing

  • edited May 2014
    Dr Gasson is one of the few without a listed number http://www.reading.ac.uk/sse/about/staff/sse-staff-all.aspx


    +44 (0) 118 378 5225 is the main number for the School of Systems Engineering (of which Dr Gasson is a researcher), maybe they would know/be able to forward your call.
  • Did anyone ever manage to get a hold of him?
  • @Ian as far as I know no one has tried to contact him apart from you. Atm we are waiting for Cassox to have some free time, hopefully at the end of August.
  • I gave it a shot but to no avail.

    Cass and I are talking about focusing hard on this project, once the NIR project is finished at the end of July. Mid to end of August should show some movement forward.
  • Yeah. We're going to be wrapping up the IR thing by then. I'm seriously excited about this.
  • How important is it to get the original details? i.e. how much effort should we put into contacting Gasson before we are wasting time and should just design our own system from scratch?
  • Very.

    Imagine that someone has built a car. You like the car. The car works. Do you a) stare really hard at the car and then try to design your own combustion vehicle from scratch, or do you b) get a diagram of the car with a parts list and then build a car?

    Which one of these projects sounds like it will have the greatest success? If you just want hacker grinder cred, sure, diy all the  way. If you want to make something that works, rt proverbial fm...
  • Hey guys I am interested in this project if you need a hand. I checked out all the publicly available stuff on Warwick's setup and have been doing my own research. It seems Warwick pretty much pulls off the shelf stuff to achieve his ends and almost all of the apparatus is exterior in his setup. If I understand you correctly, you wish to miniaturize the setup to be an entirely self contained implant. If I were engineering such a device from scratch, I would start with a cheaper array like an Atlas or Neuronexus. I would run the ribbon cable about 55mm away (to reduce noise to the probes) and then put a small hybrid fpga, a close coupled power system and a low power data system. We would need the help of a neurologist to plan and we would have to decide exactly what signals we are after. Do we want to mess with sensory input or motor output? Where do we want to intercept the nerve? We also want to add as many layers of safety as we can. By the time things start up, I could have some drawings and info 
  • To some extent, I can reconstruct the schematics from the information I've been able to comb from Warwick's papers, his book, etc, and depending on exactly what capabilities we want the implant to have we'd be modifying the design anyway -- certainly if we want to fit the whole thing in an implant.  But it would certainly help to have the originals as a starting point.

    @knal03: I appreciate the offer.  That Neuronexus company sounds interesting.  Not only are its products a lot cheaper, but some of its arrays actually provide higher resolution than the one we were looking at (not that we'd be able to use all of it anyway).  In fact, we could get a 64-channel array for about a grand.  Doing so might even mean that we could get a couple of arrays, in case one fails or multiple people want to do the experiment.

    The distance between array and circuitry is actually a bit shorter than what I was designing.  While I'm all for reducing noise, there's also the problem of infection.  If we keep the driving circuitry relatively far from the array, then if it gets infected, it's possible to notice and remove the implant before it reaches the nerve -- at least, if the surgeons use the same technique that the ones who did Warwick's implant did.  My designs currently use a 10cm long cable, FWIW.

    The fpga is actually a pretty good idea, and one that I'll definitely be thinking about in the future.  Yes, we would need the help of a neurosurgeon to do this.  Assuming we couldn't get this past a hospital's ethics board (which we probably don't have a snowball's chance in hell of doing), medical tourism would probably be our best bet.  To some extent, I do have an understanding of the kinds of signals nerves use, and we're actually after both sensory input and motor output.  Both of them more-or-less behave like sawtooth waves, particularly after applying current to the nervous system.  They also exist within a certain frequency band, but I can't remember off the top of my head (it's the same as what Bionic Tech's monitoring equipment was, though).

    Ideally, we would intercept the nerve in the same place Warwick did: on the median nerve in the arm, close to the wrist.  While a tad riskier than other nerves, it has the advantages of being relatively close to the skin at that point, and of having plenty of juicy signals to pick up.

    Yes, we certainly need to have safety features, whether it's to stop too much current from being applied, to stop the battery from exploding, to stop hackers from getting into the thing, etc.  I'd certainly love to compare designs and work with you on it.

    ~Ian
  • Some of those Neuronexus arrays look really neat, they even say they can provide arrays that last up to a year, which is more than we can expect from the blackrock array.

    Yes, if we can get them for ~1k, it seems reasonable to built more than a single device.

    Yes, medical tourism is what I assumed we'd need to do if we reach the point of implanting it.


    "battery from exploding" no no no, one of the main feature of this device is no battery. Having a battery has a number of down sides: less safe, extra size, more complex electronics, limited life of implant. Much better to just power externally when you want to run it.
  • I can very much agree on the battery part. Having no battery simplifies a lot of things. In case you really feel the need for adding one later on, you can add a battery and a transceiver circuit and place the 2 coils on top of each other. So you'd have the battery in a separate implant, relatively easy to add in a surgery.
    I'd still love to see an actual circuit for driving neurons and the waveform used. Reading them is pretty straight forward.
  • They are using a sawtooth that rises over 200usec off for 100usec and then the inverse. The freq is mostly between 1 and 3.5k
  • @knal03 current or voltage, and to what values?
  • The current is from 80uA to 100uA with 20V peak to peak. 
  • Although, voltage will change depending upon impedance. Presumably the values will have to be adjusted a bit to get nerve activation for individual body chemistry. 
  • That is a cool project, but it seems to be geared towards optogenetics,  which is a good ways from being ready for human testing. 
  • Yeah, I misspoke when I mentioned a battery.  Although I do think that having one would be advantageous in the long run, for example by giving the ability to program more complex tasks into the implant, for this initial version we don't need that.

    @ThomasEgi I can confirm the sawtooth wave picture.  It looks like that's more or less what Warwick used, and it (roughly) mimics the kind of signal that the body naturally uses to trigger action potentials.
  • I think most of that site was electrophysiology and they have some cool stuff. They are connected with a lot of good people too. They might be worth reaching out to. I found these guys on their links page. Open MEA? https://sites.google.com/site/neurorighter/

    I think this might be the same lab that did MEART maybe? (A very relevant project, btw).
  • Hi there, though I am not an electrical engineer, I am a student of electrical engineering, entering my freshman year come fall semester.  I don't know how much help I could be, but I would gladly help as much as I can.  Also, as far as using an open source license, when coding, I prefer to use the libpng license, which could theoretically work with a limited amount of reformatting and "porting" (so to say).

    I've been interested in Warwick's work since I first discovered it and I think this is a huge task, but we are hackers, after all.  Cheers and I hope it all works out.
    Nora
  • That neurorighter page is off the hook. Excellent stuff.

Sign In or Register to comment.