Voxel Based Morphometry VBM with CIVET - CoBrALab/documentation GitHub Wiki

Run CIVET with the option -VBM

In the output folder you will get a VBM folder which will have 3 minc files with 3 different tissues: CSF, GM, WM.

To analyze the VBM data, create the main dataset in R. From that dataset, add the MINC files for each subject:

Example

For gray matter

morpho_dataset_3$vbm_gm <- paste("/media/egarza/INP_MRI_Backup/projects/INP/conectome_inp/analysis/civet_JUL2017/output/",morpho_dataset_3$group,"_",morpho_dataset_3$mri_id,"/VBM/coca_",morpho_dataset_3$group,"_",morpho_dataset_3$mri_id,"_smooth_8mm_gm.mnc",sep="")

For white matter

morpho_dataset_3$vbm_wm <- paste("/media/egarza/INP_MRI_Backup/projects/INP/conectome_inp/analysis/civet_JUL2017/output/",morpho_dataset_3$group,"_",morpho_dataset_3$mri_id,"/VBM/coca_",morpho_dataset_3$group,"_",morpho_dataset_3$mri_id,"_smooth_8mm_wm.mnc",sep="")

You should now have a variable with the gray matter VBM and the white matter VBM for each subject. Make sure the files exist.

If you look at the VBM for each subject, you will notice they include the skull. We will need to focus the analysis with a mask. If you know which model CIVET used for the Normalisation, then you can use that and its brain mask.

For people in Scinet, here are the ones I used:

/project/m/mchakrav/quarantine/CIVET/1.1.12/build/Linux-x86_64/share/mni-models/mni_icbm152_t1_tal_nlin_sym_09a.mnc
/project/m/mchakrav/quarantine/CIVET/1.1.12/build/Linux-x86_64/share/mni-models/mni_icbm152_t1_tal_nlin_sym_09a_mask.mnc

I created a new folder (VBM) for all the analyses.

The mask has to be the same dimension, so we need to resample if they are not the same. You can check with mincinfo.

To resample, do:

mincresample mni_icbm152_t1_tal_nlin_sym_09a_mask.mnc -like ../output/HC_001/VBM/coca_HC_001_smooth_8mm_gm.mnc mni_icbm152_t1_tal_nlin_sym_09a_mask2.mnc

where the like file is an individual VBM file.

Load RMINC.

Linear models

You want your linear model with a mask:

VBM_gm_vs <- mincLm(vbm_gm ~ group + age + sex, morpho_dataset_3, mask = file.path(volDir, "mni_icbm152_t1_tal_nlin_sym_09a_mask2.mnc"))

Using TFCE for multiple comparisons

Make sure you have RMINC loaded and you need the TFCE command from terminal. If you are using the VM in VirtualBox, you don't need anything. If you are using your own PC, then you need to make sure Minc-Stuffs is installed. If it won't run in R-studio, you will need to run R-studio from the terminal to include the .bashrc.

VBM_gm_tfce <- mincTFCE(VBM_gm_vs, R= 5000, d= 0.1, E = 0.5, H = 2, side = "both", replace = FALSE, parallel = NULL)