Hi.
> I'm looking at the output file after running feat, specifically the
> filtered_func_data.hdr
> file, my impression from one of your emails is that that data has been demeaned
> (each voxel done individually andbesides other processing). I used your program
> avwstats -R to check the range of intensities and got 0 to 15048, the original
> data had a range of intensities from 0 to 1400. I'd like to use the demeaned
> dataset
> and also undestand what I'm looking at in the filtered_func_data.hdr file.
The filtered_func_data is the pre-stats pre-processed data, but this does
not include demeaning. However, once the stats is then run, the first
thing that the stats program film_gls does is to demean the data
internally. We do this partly so that the data never appears to stop
"looking" like real brain data, for various practical reasons (for
example, if someone were to apply their own thresholding later).
When you are using avwstats -R the range has increased because of the
global (4D, ie grand mean) intensity normalisation (which happens even if
you leave "intensity normalisation" turned off) - so this all looks ok.
If you DO actually want to demean the data yourself then I could send you
the latest version of avwmaths which would make this easy.
ttfn :)
> Thanks for your help, I really enjoy working with fsl
thanks!
|