FieldTrip

[Robert Oostenveld]

Hi Litvak,

Happy new year. Since you are asking some questions that also might be
interesting to a wider audience, I will forward this mail to the
FieldTrip discussion list as well.

On 28 Dec 2004, at 11:06, Litvak Vladimir wrote:

> Dear Robert,
>
> I've recently downloaded the FieldTrip toolbox and I have some
> questions I hope you can help me with:
>
> 1) I've been working with EEGLAB for quite some time now and I know
> you as a major contributor there as well. Why is FieldTrip developed
> separately? Are there any plans to integrate the two toolboxes in the
> future?

I get that question more often and have just entered is in the FAQ.
Please see http://www2.ru.nl/fcdonders/fieldtrip/faq.html#q7

Since both packages have a different aim, they will not be completely
merged. But on some specific parts they will be more closely
integrated. Fieldtrip and EEGLAB will share parts of their code and the
upcoming "DIPFIT 2" module of EEGLAB will be completely based on
Fieldtrip.

> 2) We have a 64 channel Quickamp EEG system and also have a
> possibility to get structural MRIs of our subjects. Would it be
> possible to use the DICS method implemented in FieldTrip for analyzing
> this data? How can I create a head model without using CTF software
> (which I understand is something commercial that comes with your MEG
> system)?

The CTF software is commercial software for MEG, and therefore not of
any use for your EEG.

FieldTrip supports various headmodels. But DICS or other beamforming
techniques are more sensitive to volume model errors than conventional
dipole fitting methods. Therefore a sperical headmodel (as for example
used in EEGLAB+DIPFIT) is not usefull. You should use a realistic
shaped headmodel, for example using the boundary element method which
is available in Fieldtrip.

However, creating a realistic headmodel for EEG data from an individual
anatomical MRI is difficult and requires quite some user interaction.
At the FC Donders Centre we have two commercial packages that we can
use for this: Curry and ASA. Therefore it has not been an issue for us
sofar to implement an open source version of a BEM headmodel
construction routine.

> 3) How can I co-register the MRI scan with electrode positions
> measured from the same subject? We didn't use capsules to mark the
> reference points but I can mark them anatomically and we have the
> measurements of the corresponding points with the electrode positions.

You can do that using a homogenous transformation matrix (see
http://www2.ru.nl/fcdonders/fieldtrip/faq.html#q5 and
http://bishopw.loni.ucla.edu/AIR5/homogenous.html). In the private
directory of FieldTrip you can find a function "headcoordinates.m" that
will help you to compute this matrix.

> 4) If there are no free tools available for creating a head model
> would it be possible for me to get at least one model for a standard
> MRI scan given our electrode positions?

In principle you could do most of this using open source tools. You
need to 1) segment the MRI, 2) create a triangulation and 3) compute
the BEM system matrix. Step 1 and 2 are possible with various open
source packages, although you probably will run into some problems with
step 2. But I do not know an open source package that can perform step
3. Perhaps I could help you to construct a single BEM model based on a
individual MRI, contact me if you are interested.

best regards,
Robert

----------------------------------------------------------------------
Robert Oostenveld, PhD
F.C. Donders Centre for Cognitive Neuroimaging
Radboud University Nijmegen
phone: +31-24-3619695
http://www.ru.nl/fcdonders/
----------------------------------------------------------------------
N.B. Starting from 1 September 2004, the University of Nijmegen has
changed its name to Radboud University Nijmegen. All web- and
email-addresses ending in ".kun.nl" should therefore be changed into
".ru.nl". Please update your address book and links.