Hi,
On 15 Feb 2008, at 19:28, Dianne Patterson wrote:
> Dear Group,
> I am characterizing white matter hyperintensities.
> "Cluster" seems to be quite useful for localizing and counting the
> wmh.
> And with the trick Steve describes below, I can get the mean value
> of each cluster and the location of the local maxima.
> I am delighted.
>
> It would be useful, though, if I could also extract other stats for
> these clusters, like range or standard deviation...
> Is there any way to pull this off?
You can use the cluster mask index image in conjunction with fslmaths
and fslstats to get range and standard deviation; use fslmaths to
extract each cluster from the index image and then use that as a mask
input to fslmaths to get the final stats.
> Finally, what does COG mean in the cluster output table? e.g., Z-
> COG X
Centre-of-gravity (intensity).
Cheers.
>
>
> Thankyou,
>
> -Dianne
>
>
> On Sun, Feb 10, 2008 at 9:09 AM, Steve Smith <[log in to unmask]>
> wrote:
> Hi,
>
> This looks fine but will not allow you to extract meaningful local
> maxima within clusters. For that, you should add:
>
> --cope=your_raw_tstat1
>
> which will 'fool' cluster into reporting local maxima in the
> 'cope' (in fact raw tstat) image as well as other summary information.
>
> Cheers.
>
>
>
>
> On 8 Feb 2008, at 09:17, Soohyun Cho wrote:
>
> > Thank you for your response.
> >
> > Yes, I used the maxc_tstat as the input to get information about the
> > cluster.
> > e.g., cluster --in=file_name --thresh=.95 --oindex=name_clusterindex
> > --olmax=name_clustermax.txt --osize=name_clustersize >
> > name_clustertable.txt --mm
> >
> > If I want to get the location of the max voxel within a cluster
> > thresholded at 1-p>.95 (corrected for cluster extent), I think you
> > are suggesting that I use raw tstat as the input?
> > In this case, how should I threshold this?
> > Can you give me an example syntax that you would use?
> >
> > Thank you,
> >
> > Soohyun
> >
> > On Feb 8, 2008, at 12:28 AM, Steve Smith wrote:
> >
> >> Hi,
> >>
> >> On 6 Feb 2008, at 20:37, Soohyun Cho wrote:
> >>
> >>> Hi,
> >>>
> >>> I have a question about using the cluster tool on randomise
> output.
> >>> The cluster table gives you the locus of a maximally active voxel.
> >>> But don't all voxels within a cluster (identified from randomise)
> >>> have the same 1-p value ?
> >>
> >> Yes, that's right, the cluster has just one p-value. However, if
> >> you used the options correctly you may have also fed in the raw
> >> tstat image into cluster as wel as the maxc image. Exactly what
> >> cluster command did you use?
> >>
> >> Cheers.
> >>
> >>> How did the cluster tool pick that Max voxel?
> >>>
> >>> What is the best way to report the loci of clusters from
> randomise ?
> >>>
> >>> Thank you in advance.
> >>>
> >>> Soohyun.
> >>>
> >>
> >>
> >>
> ---------------------------------------------------------------------------
> >> 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
> >>
> ---------------------------------------------------------------------------
> >>
> >>
> >
>
>
> ---------------------------------------------------------------------------
> 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
> ---------------------------------------------------------------------------
>
>
>
>
> --
> Dianne Patterson, Ph.D.
> [log in to unmask]
> ERP Lab
> University of Arizona
> 621-3256 (Office)
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|