Dear Muhammad,
On Wed, May 5, 2010 at 3:09 PM, MP <[log in to unmask]> wrote:
> Hi Vladimir and other SPM EEG experts,
> I am averaging my ERP data both in time and time/frequency domains and I am
> using SPM's default electrode locations for 64 (10-20) electrodes. In my
> time domain averages, the electrode locations look fine, while in my
> time-freq data, all the electrode (I am selecting 24 of the 64 electrodes
> for the TF decomposition and a frequency window of 3 - 30 Hz) seem to be
> located on the front of the head.
> The section of my script that does TF decomp is as follows:
> ...
> S = [];
> S.D = <filename>;
> S.tf.frequencies = [3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
> 24 25 26 27 28 29 30];
> S.tf.Mfactor = 5;
> S.tf.channels = [1 2 3 8 10 12 17 19 21 26 28 30 36 38 49 46 48 50 54 56 58
> 60 61 62];
> S.tf.phase = 0;
> S.tf.pow - 1;
> D = spm_eeg_tf(S);
> ...
I think I'll need a data example to understand what you mean. Could
you send me the file that you use as input to spm_eeg_tf via
yousendit?
>
> However, when I convert these TF averages to images and do stats on them, I
> see results around the right (original) electrode sites, but when i extract
> the power for that electrode and do stats outside of SPM, the results don't
> correspond to the SPM results. Is it due to mis-registration of electrode
> during the TF decomp step?
>
Here it's also not quite clear what kind of differences you get.
Please provide more details.
Vladimir
|