Hi Steve,
Thank you so much for your help with debugging. After reading your last email, I tested ICA with the ICA-FIX denoised clean NIFTI files (rfMRI_REST1_LR_hp2000_clean.nii.gz) directly downloaded from connectomedb in 10 subjects, and so far melodic has run past the point where it previously broke. So, I think you're correct that the error is from the data, specifically the way I demeaned and concatenated the files.
The commands I used to demean and concatenate the rfMRI files for one subject (100307) is below:
fslmaths /Volumes/macX/hcp/twin/subjects/100307/MNINonLinear/Results/rfMRI_REST1_LR/rfMRI_REST1_LR_hp2000_clean.nii.gz -Tmean /Volumes/macX/hcp/twin/subjects/100307/MNINonLinear/Results/rfMRI_REST1_LR/mean_rest1_LR.nii.gz
fslmaths /Volumes/macX/hcp/twin/subjects/100307/MNINonLinear/Results/rfMRI_REST1_RL/rfMRI_REST1_RL_hp2000_clean.nii.gz -Tmean /Volumes/macX/hcp/twin/subjects/100307/MNINonLinear/Results/rfMRI_REST1_RL/mean_rest1_RL.nii.gz
fslmerge -t /Volumes/macX/hcp/twin/subjects/100307/MNINonLinear/Results/concat_rest1.nii.gz /Volumes/macX/hcp/twin/subjects/100307/MNINonLinear/Results/rfMRI_REST1_LR/demean_rest1_LR.nii.gz /Volumes/macX/hcp/twin/subjects/100307/MNINonLinear/Results/rfMRI_REST1_RL/demean_rest1_RL.nii.gz
If the demeaned and concatenated file would also be of use, please let me know what platform you'd prefer me to use to share the file. Thank you again; I really appreciate your help!
Jenny
|