Hi,
On 23 Feb 2009, at 14:15, 孟春 wrote:
> Hi, I have two questions
>
> 1) When Melodic GUI carried out temporal band-pass filtering
> ( highpass filter cutoff - 100 s), the command created was:
>
> $FSL/bin/fslmaths prefiltered_func_data_intnorm -bptf 25.0 -1
> prefiltered_func_data_tempfilt
>
> I am really puzzled about the usage of fslmaths -bptf and the
> definition of sigma here. As described in help of fslmaths, 25
> should be the sigma ( in volumes). I don't know how 25 come from the
> cutoff ( 100 seconds ) while TR=2s. In my opinion, 100s means 50 TRs
> and the command should be : fslmaths -bptf 50 -1.
The 'cutoff' is defined as the FWHM of the filter, so if you ask for
100s that means 50 Trs, so the sigma, or HWHM, is 25 TRs.
> If I want to do a band-pass filtering on a 4D data (TR=2 s,
> band=0.01Hz to 0.08Hz), what parameters might be proper in "
> fslmaths data.nii.gz -bptf low_cutoff high_cutoff
> filtered_data.nii.gz " ?
> 2) In earlier version, the smooth script was " fslmaths data -kernel
> gauss sigma (in mm)". For example, FWHM=8mm, sigma will be 3.3970.
> In the current version ( FSL 4.1.0, Melodic 3.09), the smooth used
> by GUI was " susan prefiltered_func_data_thresh brightness_thresh
> sigma ..."
>
> Is it an improvement? If I want to use commandline to do the smooth,
> how to get the value of brightness_thresh? In help, this parameter
> was greater than noise level and less than the contrast of edges to
> be preserved.
The spatial extent is still the same - namely a Gaussian kernel of a
given sigma. However susan is used instead of fslmaths in order to use
susan smoothing which applies Gaussian smoothing except at the edges
of the brain, where the smoothing area is limited to being within the
brain mask - hence the edge doesn't get blurred. See the FEAT scripts
for more details on how the susan parameters are determined.
Cheers.
>
> Thanks in advance!
> --
> Chun Meng
> NeuroImage Computing group,
> State Key Laboratory of Cognitive Neuroscience and Learning,
> Beijing Normal University,
> Beijing 100875, China
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|