DTI Preprocessing steps (for TBSS)

Charlotte Rae 2012


copy_series_dcm2nii
Make a folder for each subject in the study directory
Copy the DTI Series from the CBU mridata to subject folder
dcm2nii in the copied Series folder

eddy_correct_data
Move bvals, bvecs and data.nii.gz from the subject_folder/Series*/ to subject_folder/, and rename it something sensible
Eddy correct the data, and put "ec" at the end of the filename

dti_motion_data
Calculate and get pictures of how much subjects moved, using Mark Jenkinson's "dti_motion" script

--Might want to double check movement is not excessive (roughly equal to voxel size, e.g. 2mm on standard CBU diffusion sequence). View ec_trans.png or ec_rot.png in a linux graphics viewer, e.g. Eye of Gnome (type eog at command line).

brain_mask
Extract the b0 image (usually first in the volume series - check) and call it nodif
Bet the nodif and call it nodif_brain. Also apply option -m and get a nodif_brain_mask

--Might want to double check that the nodif_brain_mask looks ok: brain-shaped and binary

dtifit_data
Fit the tensors, using the data_ec.nii.gz, plus bvals, bvecs, and a nodif_brain_mask

--Might want to check at this point that the FA image looks sensible (corpus callosum red, cingulum bundle green, corticospinal tract blue). Start FSLview, and open dti_V1. Add FA as an overlay. While V1 is highlighted, click blue "i" box, and from "DTI display options" choose "RGB" and "Modulate by FA".

copy_FA
Copy each subject's FA image to the TBSS analysis directory, ready to start TBSS (can be easily altered to also copy MD images etc)
