import bva time-freq data

[Robert Oostenveld]

On 26 Jul 2006, at 17:28, Stephan Bickel, Anatomisches Inst. wrote:
> thank you very much for your response and for fixing the bug.
> Reading in the trial information works fine now. Unfortunately I
> encounter another problem when I want to read in the data with
> preprocessing. I wonder if it is yet possible to read in time-
> frequency data from bva-dat files?

...

> Error in ==> fieldtrip-20060725\private\read_brainvision_eeg at 65
>     dat(line,:) = str2num(str);
>
> Error in ==> fieldtrip-20060725\private\read_data at 239
>     dat = read_brainvision_eeg(filename, hdr, begsample, endsample);

Hi Stephan,

I have no idea how time-frequency data is stored in the BVA *.dat
files. Even if the low-level reading function were to support it, the
PREPROCESSING function at least would not be able to deal with it,
since the data structure returned by preprocessing does not allow for
3D data (channelXtimeXfrequency). That type of data would have to be
stored in a structure that is compatible with the output of
FREQANALYSIS. The preprocessing function is made for raw continuous
(or trial-based) data, and it happens to work as well for averaged
ERP data, although it is not specifically made for that.

For other external software (BESA, EEGLAB, LORETA) I have created
xxx2fieldtrip functions, i.e. functions that import already processed
data and format it into the fieldtrip structure that is most
compatible with it. In this case I can imagine a
BRAINVISION2FIELDTRIP function that would read your dat file and
return a "freq" structure. Besides making a brainvision2fieldtrip
function, it requires making either the read_brainvision_eeg smarter
(or probably making a seperate read_brainvision_tfr function) and
improving the detection of teh content of the BVA *.dat file. Please
have a look in besa2fieldtrip to see how I approach it there.

best regards,
Robert