Dear Charline,
On Thu, Feb 2, 2012 at 11:34 AM, Urbain Charline <[log in to unmask]> wrote:
> My first question is: when I used the code that you provided in a
> previous mail (31/12/2012, see attachment 1) to extract the source
> waveform in a particular MNI coordinate (28, 14 -28), it gives me the
> temporal sequence on the interest window (650-900, see attachment 2),
> but I would have preferred to plot results on all the epoch (0-900ms),
> is that possible?
This is not a valid thing to do since the solution is only correct for
the time window for which it was computed. If for instance there are
strong activations outside this time window from sources that are not
present in the solution, these activations will be incorrectly
projected to sources that are there so the waveforms will be wrong. If
you want to get waveforms for a longer epoch you should compute the
solution for that epoch.
>
> Secondly, I found this MNI coordinate of interest (28, 14 -28) at the
> group level (by producing separate source reconstructions for each
> participant. These reconstructions were then analyzed by
> proportionally scaling the results and submitting the reconstructions
> to the general linear models and contrasts of interest). I was
> wondering what should be the best way to represent the sourcewave of
> this specific result/activation? Should I reconstruct the grand
> average and plot it sourcewave ?
I don't think you can get a waveform that will be exactly analogous to
what you did with the images because you used scaling. I would get
waveforms from the individual solutions you used for the stats and try
to see something parallel to the effect that you found by plotting
them together.
Vladimir
|