Hi,
> 1. What exactly does the 1st eigenvariate represent? The VOI time course
> that SPM outputs is on the course of 150-200 (units??), while the
> adjusted and fitted responses that the plot function gives are on a much
> smaller scale. How is the time course data different from the scaled
> data and is the 1st eigenvariate a good indicator of signal changes
> maintaining the structure of the raw variance in the data?
I'm afraid the meaning of the 1st eigenvariate is rather difficult to
explain well in a short space; in general for FMRI signals it will be
close to the mean time course. The 150-200 (units??) are in the same
units as all SPM FMRI signal - the list will have some useful hints on
this. In brief, the units are relative to a whole brain mean signal
of 100.
> 2. How would I go about calculating percent change in signal using the
> time course data (without contrasts). Is it as simple as
> Y = ((Y - mean(Y))./mean(Y))*100 ??
If you just want the percent deviation of the signal from the mean
signal, then yes.
> 3. Does it makes sense to simply take the mean of the time courses to
> represent the average group time course?
I suppose so...
Best,
Matthew
|