FW from FSL

Hi - easy - use the cluster *index* output option instead

--oindex

Cheers.

On 29 Mar 2011, at 19:51, Michael B wrote:

Hi there,

I have run TBSS and have found several clusters of significant voxels. My goal is to extract information from these clusters individually, such that for each cluster, I have an FA value (derived from all_FA.nii.gz), a t-statistic (derived from *_tstat*.nii.gz), and coordinates. To achieve this goal, I created an image that labelled the clusters according to size, using the cluster command:

cluster --in=*_tfce_corrp_tstat1 --thresh=0.95 --osize=*_tfce_corrp_tstat1_size

Subsequently, I thresholded this "size image," such that I created a mask for each cluster, using the fslmaths command:

fslmaths *_tfce_corrp_tstat1_size -thr (size of cluster 1) -uthr (size of cluster 1) -bin *_tfce_corrp_tstat1_size_mask1
fslmaths *_tfce_corrp_tstat1_size -thr (size of cluster 2) -uthr (size of cluster 2) -bin *_tfce_corrp_tstat1_size_mask2
fslmaths *_tfce_corrp_tstat1_size -thr (size of cluster 3) -uthr (size of cluster 3) -bin *_tfce_corrp_tstat1_size_mask3
etc.

In this way, it became possible to extract information from the clusters individually. However, this approach falls apart when 2 or more separate clusters have the same size. In this case, the mask includes 2 or more clusters and the information extracted is from several clusters collectively, not from the clusters individually. How might I solve this problem?

I would dearly appreciate some guidance on this issue!

Thank you,
Michael

---------------------------------------------------------------------------
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)
steve@fmrib.ox.ac.uk    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------