| T O P I C R E V I E W |
| mullette-gillman |
Posted - Jun 10 2010 : 5:06:40 PM
Hi there,
I am trying to use the FSL 'cluster' command to find peak coordinates within a large cluster of activation. With this, I should be able to find multiple local maxima within a single activated cluster. I've done this previously (old cluster, prior version of FSL), but it isn't working now. The base command (from qinteract),
cluster --in=thresh_zstat1.nii.gz --thresh=2.3 --mm --num=10
should take in the zstat image, and output 10 local minima for each cluster. Instead, it is only outputting 1, and the --num=10 (or -n 10, alternatively) argument is being ignored.
Anyone know what is going wrong here?
I know that I can raise the threshold to break the large cluster (i.e., set --thresh=2.3 to higher values) and this will give me the maxima of each produced cluster, but I don't like that solution.
Thanks,
O'Dhaniel |
| 5 L A T E S T R E P L I E S (Newest First) |
| mullette-gillman |
Posted - Jun 15 2010 : 11:58:17 PM
Thanks Dave,
That also worked for me. The connectivity term is optional, the key part is the --olmax=filename term
Thanks!
|
| dvsmith |
Posted - Jun 15 2010 : 5:47:53 PM
try this instead:
cluster -i thresh_zstat1 -t 2.3 --connectivity=26 -n 10 --mm --olmax=lmax_zstat1_std.txt
that works for me... |
| mullette-gillman |
Posted - Jun 13 2010 : 10:25:27 PM
Hi Chris,
My notes were that it worked on the thresh_zstat.nii.gz. I tried it also on the zstat1.nii.gz, and it produced the same output of 1 maxima per cluster.
Thanks though! |
| petty |
Posted - Jun 11 2010 : 2:08:10 PM
should you be running on the non-thresholded zstat1 image instead? |
| mullette-gillman |
Posted - Jun 10 2010 : 5:42:52 PM
Noticed an error in what I wrote:
"should take in the zstat image, and output 10 local minima for each cluster."
should say
"should take in the zstat image, and output 10 local maxima for each cluster."
|
|
|
