DTI Presentation_Final_11.1.2010

Report
Diffusion Tensor Imaging:
The Nitty Gritty
Brought to you by: Meenal and Erica
November 2, 2010
Terms
• FA: Fractional Anisotropy
– The eccentricity of the ellipsoid
• MD: Mean Diffusivity
– The mean of the three principle eigenvectors
• DTI Terminology Reference:
http://ccn.ucla.edu/~jbrown/dti.html
DTI Step 1: Convert Your Data
• Convert the DTI DICOM files to NIFTI format
DTI Step 1.A
• Change your FSL Output Type
– Cd: /usr/local/fsl-4.1.4/etc/fslconf
– Run: sudo nano fsl.csh
– Change FSLOUTPUTTYPE TO NIFTI
– Control X
– Save? “y”
– Relaunch the terminal and check the change
DTI Step 1.B
• Edit script as needed
– Script:
/Volumes/sasquatch/DataManagement/Workspace/P
rocessing/MatlabCode/”MY”Matlab/main_dtipreproc
ess.m
• Make sure your account system preferences are
set to /bin/tcsh
• Open matlab
• Run: main_dtipreprocess.m
DTI Step 2
• Move all folders
– done, dti,ec,fsl, medinria, ninfo.txt,reg, vc and vl
• From:
– /StudyName/Original_Data/nii/subjid/dti
• To:
– /StudyName/Original_Data/DTI/subjid/dti/
DTI Step 3
• Rename files in:
– StudyName/Original_Data/DTI/subj/dti/fsl/
– From: *rec*
– To: *$subjID*
DTI Step 4
• Create folders:
– Studyname/DTI_Analysis
– Studyname/DTI_Analysis/StudyName_MD
– Studyname/DTI_Analysis/StudyName_MD/origd
ata
– Studyname/DTI_Analysis/StudyName_FA
– Studyname/DTI_Analysis/StudyName_FA/FA
DTI Step 5
• Copy data
– Files ending in *FA*
– From: StudyName/Original_Data/DTI/subj/dti/fsl
– To: Studyname/DTI_Analysis/StudyName_FA/FA
– Files ending in *MD*
– From: StudyName/Original_Data/DTI/subj/dti/fsl
– To:
Studyname/DTI_Analysis/StudyName_MD/origd
ata
DTI Step 6
• Run Tract Based Spacial Statistics using the FSL
Package
– CD into DTI_Analysis/StudyName_FA/
– Copy dti_tbss1-3.sh
– Run script
– Note: Errors may occur here if fsl is not set up
correctly
TBSS
• To Learn More
– http://fsl.fmrib.ox.ac.uk/fsl/tbss/
• preprocessing - create FA images from your diffusion study data
• tbss_1_preproc - prepare your FA data in your TBSS working
directory in the right format
• tbss_2_reg - apply nonlinear registration of all FA images into
standard space
• tbss_3_postreg - create the mean FA image and skeletonise it
• tbss_4_prestats - project all subjects' FA data onto the mean FA
skeleton
• stats (e.g., randomise) - feed the 4D projected FA data into GLM
modelling and thresholding in order to find voxels which correlate
with your model.
DTI Step 6
• Check mean_FA
DTI Step: 7
• Create sublist in DTI_Documentation
• Create folders:
– DTI_Analysis/WM2DTI2Template
– DTI_Analysis/GM2DTI2Template
DTI Step 8
• Separate mean_FA and mean_MD into gray and white
matter
– Set up and Run: dti_combined_StudyName.sh
– This script:
•
•
•
•
#Step 1: Reorient the hires to Match the DTI
#Step 2: Create Average Bo
#Step 3: Register hires to Average Bo
#Step 4: Apply warp of hires in DTI space and MD to MNI152
– # Threshold the MD Map > 0
– # Apply the warp (Subject FA to FMRIB_FA template) on MD map
•
•
•
•
# Step 5: Segment the hires in MNI152 space
# Step 6: Threshold just the WM and GM (Whole-brain)
# Step 9: Multiple registered FA map and mask (WM and GM)
# Step 10: Calculate the Mean FA for WM/ mean MD for WM and GM
DTI Results
• Bo+12 directions
– FAtarget_WM
– MDtarget_WM
– MDtarget_GM

similar documents