neurofeedback with old amplifiers / open source?
paulbrennan
Victoria, Canada
HI there,
I am just learning about NFB (as a psychologist), maybe with a view to eventually becoming board certified. At this point I would only practice on myself, and even then with some caution. I'm interested in the open source hardware/software options here as a way to learn but at some point I'm going to hit a snag that they are not approved Class II devices and it may be difficult to get other NFB professionals to mentor me or, which is required. However I see a lot of older regulated amplifiers on eBay at low cost. I wonder, is the interface/drivers/etc usually proprietary in these cases, or is there a standard protocol that I could use with at least some of these devices. In other words, could I probably get an old amplifier to work with OpenBCI, given average tech skills, at least as a starting point?
Another way to look at it is from the software angle.... If I got a subscription to one of the clinical software applications like EEGER, could I then get more or less any hardware to work with it or is it highly restrictive? I see I could get a $1000 amplifier from Pocket Neurobics - not regulated but at least it's well known - that integrates with EEGER. Any advice would be really welcome!
Kind regards, Paul
Comments
Now that I read it back, this post is probably in the wrong place. I guess I'm really just looking for any perspectives on the interoperability of NFB systems, both clinical and non-clinical.... Gotta start somewhere!
Paul, hi.
Most neurofeedback apps that run on desktops / laptops (vs mobile), use amplifiers with proprietary interfaces. For example the Pocket-Neurobics Q-Wiz / *-Wiz family. But Pocket-Neurobics DOES share their interface details with certain app developers, such as EEGer and Bioexplorer. So it would be challenging to for example, interface OpenBCI to EEGer, (in PN emulation mode) without such details. Additionally, there is the issue of sample rate differences between amplifiers. For example 256 Hz vs 250 Hz vs 200 Hz, etc.
One highly regarded neurofeedback app that DOES talk to OpenBCI, is BioEra. BioEra is the underlying low level layer that forms the basis of such commercial packages as Cygnet, LENS, etc. And just this year, Pete Van Deusen's Brain-Trainer package is now running with BioEra. (Pete has supported Bioexplorer amps for many years.)
https://brain-trainer.com/
http://bioera.net/
Regards, William
Related to the previous comment, if you are looking at a particular old amplifier on eBay, check with the list of supported amps on BioEra and Bioexplorer websites. If they support it, then Pete's system should be able to talk to it.
http://www.proatech.com/be/manual.html#_Toc39305931
http://www.cyberevolution.com/hardware.htm
[oddly the Bioexplorer site does not list the PN *-Wiz devices, but they are supported; as well as older PN devices.]
The other thing to note about Bioexplorer, is that tech support or upgrades have largely been abandoned. Though many of Pete's past customers are still using it.
EEGer amplifiers are listed in the Tech manual and bottom right link on this page:
http://support.eeger.com/
It's possible EEGer could be encouraged to support the OpenBCI Cyton and Ganglion, especially now that the Brainflow libraries are available:
https://brainflow.readthedocs.io/en/stable/
Two other apps that can do neurofeedback with OpenBCI are:
https://www.neuromore.com/
http://www.shifz.org/brainbay/
But these do not come with protocol suites like Pete's or EEGer. However the element blocks for signal processing are there to construct arbitrary reward / inhibit designs.
Thanks William, that's extremely thorough and helpful. I'll probably take a look at BioEra as a starting point. "BioEra is the underlying low level layer that forms the basis of such commercial packages as Cygnet, LENS, etc." - Which is interesting b/c I think BioEra is not FDA approved device whereas LENS etc. are! It's a strange system.
For right now, I have a Muse with an extra electrode and a Myndlift provider subscription. I know this is not regarded as a clinical grade system by most NFB people - though I'm not exactly clear on what the actual technical limitation is other than being single channel (plus reference). It's not FDA approved but then neither are a lot of systems out there apparently. What's your take on the FDA approval aspect - is it just a label?
I'm not clear what advantage "FDA approval" offers for neurofeedback applications. There are many many (consumer level) apps now existing in this field. Quite a few that work with mobile devices (such as Muse, Emotiv, etc.). Which was not the case a decade ago. If you are looking for BCIA certification, you'd be best going with whatever the instructor is training with. But hundreds of home trainers are using Pete's system, which has no official recognition.
https://www.google.com/search?q=does+fda+regulate+neurofeedback
This might have been relevant when neurofeedback was only offered by "licensed practitioners". But that era seems to be less and less relevant.
Yes good observation. I don't think it can matter much at this point. I'm in Canada but I don't think it's much different.
Another Canadian system that has had wide recognition and success is NeurOptimal. Which I do not believe has any FDA status, other than being a general "wellness device".
https://neuroptimal.com/
https://neuroptimal.com/wp-content/uploads/2018/12/FDA-letter.pdf
https://neuroptimal.com/fda-guidelines-statements/
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfStandards/detail.cfm?standard__identification_no=32789
Yes a lot of people around here are offering services with Neuroptimal. In fact it looks from your attachment that they are out of Victoria, my town! To me, it seems a bit overpriced (equipment plus services charges) but I guess I can't have it both ways.
Here is some background with a bit of explanation,
https://neuroptimal.com/blog/history-of-neuroptimal/
There used to be some old free videos / audios of Val Brown (developer) on the web. Just searching now, don't see those. But there are some old recordings for purchase:
https://www.futurehealth.org/author/productauthor418.html
But as he explained it at the time: NO is detecting 'turbulence' at C3 C4, which Brown equates with the central nervous system "changing state" to a different thought stream. During the session, NO creates a brief click or audio interruption at these state changing moments. Thus during the session the client gains more awareness about when he is drifting off into possibly old "tape loops" of mental patterns. By training to unconsciously recognize these state changes, the client gains more ability to stay in the present moment and not get pulled into old (possibly dysfunctional) patterns.
It seems like a very interesting model and I'd like to try it. On the other hand, like the Muse, it sort of does one thing, which gives me pause - I wonder what you do if this model doesn't seem to producing the expected results? Right now I'm reading the Sebern Fisher book which suggests trying a number of strategies and tweaking them based on the client's response, history, etc., so it's kind of at the other end of things. Myndlift offers some preconfigured protocols, but you can also create your own. I'm finding it a good way to learn and the cost is reasonable.
You could say the same of Cygnet and LENS. They only do one type of neurofeedback, and not adaptable to others. Yet substantial numbers of practitioners decide to specialize in one of these 'different' types of NFB. Both Cygnet (ILF) and Neuroptimal could be considered 'non-volitional' feedback paradigms. They don't push the brain in a particular direction, but just give moment to moment feedback about what the brain is doing. Then the central nervous system witnesses itself, and self corrects.
Yet the underlying platform (BioEra) is capable of all of these. Because it is a general purpose signal processing framework. Including band based reward-inhibit protocols, such as those used by Pete and many other systems.
Sebern's book is a landmark.
https://www.sebernfisher.com/
William I don't really understand Pete's system yet but it seems like a layer on top of BioEra, BioExplorer. Which I also haven't used. But do you think that layer is really necessary? Is it just a matter of a better GUI?
Also I do wonder if something like Myndlift/Muse - if it were just a little bit better established - would maybe outdo all of these, so therefore it's a tricky time to invest in these desktop systems...
No, Pete's Brain-Trainer is a complete package with assessment and automatically generated / suggested protocols.
ALL the commercial clinical / home-trainer NFB systems (using BioEra: LENS, Cygnet, Brain-Trainer), are layered on top of BioEra as a signal processing framework / toolkit. Pete's Bioexplorer based package is open source. Whereas he closed it for BioEra. Very sad to hear that, because with Bioexplorer you could learn from and tweak the designs.
ALL of these use GUI features build upon BioEra's / Bioexplorer's base features of signal processing, GUI, and feedback.
The Myndlift sounds like a good direction. But very limited in sensor placement. PN equipment does not have this issue.