Hi Mary
new bandpass filter removes the mean (as a real bandpass filter
should). You can check this old thread "Fslmaths bptf differences in
5.0.7 - demean"
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1408&L=fsl&P=R73559&1=fsl&9=A&J=on&d=No+Match%3BMatch%3BMatches&z=4
best
Enrico
On 23 February 2015 at 01:20, Mary Askren <[log in to unmask]> wrote:
> In previous versions of FSL (5.0.6 and 4.1), we were using the -bptf flag in fslmaths to apply a high pass filter to our fMRI data. However, using the same command in FSL 5.0.7 (from NeuroDebian) yields unexpected output that is no longer looks brain-like. Has anyone else experienced this? Is there a reason for this change?
>
> We have been using a command like fslmaths input.nii.gz -bptf 33 -1 filteredoutput.nii.gz on fMRI data with TR=2 seconds, but the output data look similar with a wide range of highpass sigma values (e.g., 2, 100). I can provide example data if that would be helpful.
|