[Neurodebian-users] real-time EEG viewer

Nicolas Bourdaud nicolas.bourdaud at gmail.com
Mon Jan 30 12:12:53 UTC 2012


Hi,

I missed these emails at the beginning of the month, but let me replied
any way.

On Thu, Jan 05, 2012 at 14:11:41PM -0700, Michael Hanke wrote:
> On Wed, Jan 04, 2012 at 04:05:49PM -0700, Carl Turner wrote:
>> Yes, I'd like to interface the AvatarEEG acquisition hardware via
>> Bluetooth to open source software.
>>
>> Thanks for your good suggestions. I had not come across trellis-neuro,
>> relacs or rtxi. So far I have looked at BrainBay and Edfbrowser.
>>
>> A package that is popular with the neuroscience community would be
>> ideal.
>
> We are also working with these guys
>
> http://cnbi.epfl.ch/
>
> to package their software stack. Base libs are already there, the rest
> is coming.

We have indeed a solution that corresponds to your requirement. In our
lab, we have developed a set of libraries which handle, each of them, a
specific part of the processing pipeline (signal acquisition, filtering,
recording, signal display...) and on top of those libs, some programs
that do the integration (eegview is one of them and do exactly what you
need: display signal from a device in a scope and record it if you want).

	Each of the libraries has been designed to be completely independent
from each other, and their design follows the unix way: Do only one
thing and do it well. We have done this way, because we were not happy
with the existing "popular" solutions which implement everything as a
framework which makes extremely difficult to use a subpart of it
independently from the rest.

	For your problem, I think you may be interested in "eegdev" which is a
library that provides a flexible and device independent interface to
biosignal acquisition devices (in particular EEG). The design is very
modular: support for a particular device is achieved by writing a small
plugin, while many of the common functionalities (data casting, channel
selection, providing metadata, error reporting...) are already handled
by the core library. The library will be very soon in NeuroDebian and
Debian proper (I think now it is a matter of days). Also there is
already the implementation for a bluetooth based EEG device, so
implementing a new one, can be inspired by it.

	
	Why do I advocate for it? Simply because the
neuroscience/neuroengineering community really lacks of interoperable
tools and should stop wasting resource through duplicated work. Too many
frameworks reimplement the support for the same hardware. All of this
also means less tested acquisition code (which is a critical part). Also
I have seen sometime some device supported by a framework and not by
other: why?

	So what I call for is to unit the effort and to create an acquisition
library that could be reused by any framework we know. So that if a
plugin exists for a device, the device can be used automatically by the
framework. I know that some attempt have been done by creating a network
protocol for this, but in my opinion, it adds unacceptable complexity
and overhead. The solution must be flexible but also light and fast.

	I hope some other people think like me and we can joint our work to
achieve this. eegdev is just my proposal for this and I think it covers
quite well the need. But it can be amended if needed.


	Cheers,

Nicolas

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 900 bytes
Desc: OpenPGP digital signature
URL: <http://lists.alioth.debian.org/pipermail/neurodebian-users/attachments/20120130/027e41c5/attachment.pgp>


More information about the Neurodebian-users mailing list