Dear FSL experts,
I received no answer for the message below.
Can somebody help me in understanding if this is a problem or a normal outcome?
I have checked all registrations and they seem all OK, but there is no perfect overlapping between the registered-EPI images among them and between the same images and the standard MNI.
Is this normal or is it a problem for running melodic?
thank you
Angela
Dear experts
I am trying to use MELODIC by using scripts.
I have performed all the preprocessing (bet, motion correction, smoothing, temporal filter, detrending) and finally registration using the following commands:
flirt -ref highres -in rest -out rest2highres -omat rest2highres.mat -cost corratio -dof 6 -interp trilinear
convert_xfm -inverse -omat highres2rest.mat rest2highres.mat
flirt -ref standard -in highres -out highres2standard -omat highres2standard.mat -cost corratio -searchcost corratio -dof 12 -interp trilinear
convert_xfm -inverse -omat standard2highres.mat highres2standard.mat
convert_xfm -omat rest2standard.mat -concat highres2standard.mat rest2highres.mat
flirt -ref standard -in preprocessed_rest -out rest2standard -applyxfm -init rest2standard.mat -interp trilinear
then I run melodic with this command
melodic -i subject_melodic.txt -o groupmelodic.ica -v --nobet --bgthreshold=3 --tr=2.035 --report --guireport=../../report.html --mmthresh=0.5 --Ostats -a concat
Everything seemed right, but at the end I realised that the files thresh_zstatX.nii.gz and the mask.nii.gz generated by melodic are bigger than the MNI standard. What is my mistake? Do I need to specify a mask when running melodic? Or is it wrong my registration procedure?
Thank you!
Angela
|