[Mne_analysis] Getting an evoked file in matlab

Matti Hamalainen msh at nmr.mgh.harvard.edu
Fri Apr 3 17:27:35 EDT 2009
Search archives:

Hi Andres,

On Apr 3, 2009, at 4:51 PM, Andres Felipe Salazar Gomez wrote:

> Hi MNE users,
>
> I want to create my own evoked file using not an eeg average but a  
> chunk
> of it. I want to do this to analyze the localization of a signal that
> appears at a specific time in the eeg (not related to stimuli).  
> Therefore,
> I will take a window of eeg containing this signal.
>
> To use MNE and find the sources in the cortex, I need this  
> "averaged" file
> but I am having problems to get it.
>
> I want to take a piece (200 ms) of raw data. I loaded the raw file in
> matlab running the fiff_setup_read_raw and cut the chunk of interest  
> using
> fiff_read_raw_segment. I got a (samples x channel) file but I had  
> problems
> saving it as an averaged or evoked file. I tried using the
> fiff_write_evoked command but I do not have the data.info piece nor  
> the
> tree structure.
>
> With the raw data I got a regular averaged of some triggers and after
> loading this file in matlab I replaced the data.evoked.evoked file  
> by my
> chunk of data and changed the data.info.filename for the new filename.
> Doing this I was able to create the data structure of my "averaged"  
> file
> with data.info and data.evoked. When I run fiff_write_evoked it
> successfully creates a file.  Nevertheless, it seems this file will  
> not be
> useful since I can not load it as an evoked file in mne_browse_raw. It
> prompts (Incorrect number of samples in data matrix).

Here is what you need to do:

1. Do *not* change the filename field to the new filename. This field  
should be the original filename and is used to copy additional  
information from the original file to the destination file. It is  
enough to define the output file name as a parameter to  
fiff_write_evoked. This has no bearing to your problem, though.

2. You need to change the first and last members of the evoked data so  
that last - first + 1 = size(epochs,2). This defines the time scale of  
the data. If the data segment is not related to a stimulus, you may  
just want to set first = 0 and last = size(epochs,2) - 1. You can also  
set first and last to correspond to the sample numbers in the raw data  
file.

If you just want to analyze raw data segments in mne_analyze, you can  
also make your own events in mne_browse_raw, create an event file, and  
read the raw data directly in mne_analyze.

I hope this helps,
Matti




-------------

Matti Hamalainen, Ph.D.
Athinoula A. Martinos Center for Biomedical Imaging
Massachusetts General Hospital
Building 149, 13th Street, Mailcode 149-2301
Charlestown, MA 02129-2060
USA

e-mail          msh at nmr.mgh.harvard.edu
Tel             +1 617 726 0323
FAX             +1 617 726 7422



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20090403/4a464edd/attachment.html 


More information about the Mne_analysis mailing list