changed default reading functions for CTF

[Robert Oostenveld]

Hi Tom,

On 30 Sep 2008, at 23:09, Tom Holroyd wrote:
> Two questions. 1) We record data with 3rd gradient turned on by
> default.
> That is, the MEG channels have already had the background noise
> subtracted. So our datasets contain only balanced data. No further
> processing is required. Does your code take care of this case?

Hmmm, good question... I cannot immediately answer confirmative. The
data that is read will be balanced, that won't be the problem. What I
don't immediately oversee is whether the leadfield computation will
match the balanced data. In the old case it was sure not to match the
data. In the new case the reading routine should take care of that.

I don't have any 3rd order data lying around right now, so would you
please check the following
   data = preprocessing(cfg)  % with your standard config
and then look into the field
   data.grad
There should be a
   data.grad.balance
structure and
   data.grad.balance.current
string. In your case current should be G3BR.

Subsequently please check the weighing of the gradiometer coils to
channels using
   imagesc(data.grad.tra)
You should more or less see two identity matrices concatenated with
some small non-zero numbers in the rightmost part of thaht figure.
This matrix is used in the forward computation: first the field at
every individual coil is computed (keeping the coil orientation in
mind), and then the "coils" are combined to form "channels". The two
identity matrices represent the bottom and top coil, and the small
numbers in the rightmost part represent the small corrections to the
forward model required to mimic the balancing (in which also a small
part of the dipole field is subtracted from the data).

Please let us know whether those two fields are as I expect them to
be. Perhaps you could also send me (not through this email list!) a
3rd gradient dataset (preverably averaged, i.e. small MB size) for
your system, so that I can use that for testing.  You can use ftp.fcdonders.nl/pub/incoming
  or something like yousendit.com.

> 2) Why is
> the code closed source?

We discussed the release of the code in the joined context of SPM8 and
Fieldtrip with the CTF support. See below for their statement

>> ---------- Forwarded message ----------
>> From: MEG Customer Support <support at megservice.com>
>> Date: Thu, Apr 24, 2008 at 1:00 AM
>> Subject: RE: MATLAB Import/Export Routines
>> To: David Bradbury <d.bradbury at fil.ion.ucl.ac.uk>, MEG Customer
>> Support <support at megservice.com>
>> Cc: litvak.vladimir at gmail.com
>>
>>
>> Dear David (& Vladimir),
>>
>> We're most glad to hear that the issue with the stim codes may have
>> been
>> resolved.  As you say please keep us informed if it is not.
>>
>> Thank you very much for your request regarding the VSM MatLab codes.
>> Because we consider them to be covered under the systems' general
>> non-disclosure agreements you are free to use these algorithms
>> internally (i.e. by all MEG users at UCL and F.C. Donders Centre)
>> without restriction however we would like to request that you only
>> distribute these functions external to your labs in a pre-compiled
>> form.
>> Thank you for your consideration in this matter.
>>
>> With best regards,
>>
>> Brent


So I have the m-files but am not allowed to distribute them, hence the
p-files in the fieldtrip/external/ctf directory (and in spm8/external/
ctf). I guess that you as CTF user should be able to get your own copy
of the m-files from CTF support.

best regards,
Robert





----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.