A few questions regarding getting cluster info from randomise output
In the standard cluster output table “Voxels” = the number of voxels in the cluster.
In the randomise user guide segment “Getting Cluster and Peak Information from Randomise Output” [ here is the link http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Randomise/UserGuide#Getting_Cluster_and_Peak_Information_from_Randomise_Output ] fslmaths is used to have the tfce_corrp_tstat1 serve as in index or mask for the raw tstat1 . The output, listed in the user guide as grot_thresh_tstat1 then serves as mcp-corrected input for the cluster call (cluster --in=grot_thresh_tstat1 --thresh=0.0001 --oindex=grot_cluster_index --olmax=grot_lmax.txt --osize=grot_cluster_size)
While the cluster call output lists all clusters that survive, such clusters are of all sizes. I understand that the - -osize output from the cluster call can be further filtered.
To restrict significant clusters to only those with at least >=10 voxels, would this be correct?
fslmaths grot_cluster_size –thr 10 grot_cluster_size10
(-bin command would seem redundant here given the raw tstat1 was masked by the tfce_corrp_stat1 and binarized initially)
The newly filtered cluster call could then be as follows, correct?
Cluster –in=grot_cluster_size10 –-thresh=0.0001 -–oindex= grot_cluster_index10 --olmax=grot_l0.txt --osize=grot_cluster_size10 –mm
Also, in the standard cluster output the “Voxels” column refers to the number of voxels in the cluster. As such, it would seem the sum of the Voxels column would be another indicator of total significant clusters, here those clusters that are >=10 voxels. Is this correct?
Finally, would it be possible to use FAST in conjunction with a mask of corrected significant output (e.g. grot_cluster_size10, as above) from tbss_sym to arrive at a percentage of L>R and R>L of total white matter?
|