Skip to content

Latest commit

 

History

History
74 lines (47 loc) · 2.72 KB

KUL_T1T2FLAIRMTR_ratio.md

File metadata and controls

74 lines (47 loc) · 2.72 KB
Raw

KUL_T1T2FLAIRMTR_ratio

Purpose

Computes a T1w/T2w, T1w/FLAIR and MTC (magnetisation transfer contrast) ratio, using BIDS organised data. The full methodology is described in "T1w/FLAIR ratio standardization as a myelin marker in MS patients, by S Cappelle, D Pareto, S Sunaert, I Smets, A Laenen, B Dubois, Ph Demaerel" - Article link

Required structural input data

Required are data organised in BIDS format:

  • a T1w image without contrast
  • a FLAIR and/or a T2w image
  • a MTI image pair (optional)

Usage

A typical command for running is:

KUL_T1T2FLAIRMTR_ratio.sh -a -n 10 -f 1 -m

You need to be in main directory to run this command. This is also where the BIDS folder resides.

For info about the command just run:

KUL_T1T2FLAIRMTR_ratio.sh

Output in the folder T1T2FLAIRMTR_ratio

Images starting with sub-xxx-ses-yyy_:

  • _T1w.nii.gz: the input T1w, converted into 1x1x1 mm, biascorrected
  • _T2w_reg2T1w.nii.gz: the input T2w, converted into 1x1x1 mm, biascorrected and rigidly registered the the T1w
  • _FLAIR_reg2T1w.nii.gz: the input FLAIR, converted into 1x1x1 mm, biascorrected and rigidly registered the the T1w

Further output is in 1x1x1 mm voxel dimensions.

T1w/T2w, T1w/FLAIR ratio-maps starting with sub-xxx-ses-yyy_:

  • calib-none.nii.gz: ratio without calibration (raw T1w/T2w or T1w/FLAIR ratio)
  • calib-lin.nii.gz: LINEAR histogram matching using eye/muscle tissue (Ganzetti et al. 2014 alike)
  • calib-nonlin.nii.gz: NONLINEAR histogram matching using eye/muscle tissue (Ganzetti et al. 2014 alike)
  • calb-nonlin2.nii.gz: NONLINEAR histogram matching according to Cappelle et al. 2022
  • calb-nonlin3.nii.gz: NONLINEAR histogram matching on brain tissue, excluding white matter lesions (from samseg)

MTR ratio maps:

  • ratio-MTC.nii.gz

ROIs:

  • rois/sub-x_ses-x_MSLesion: a binary image containing the MS lesions as identified by Freesurfer Samseg

mask:

  • in folder masks one can inspect the eye/muscle, brain, background, etc masks.

Note:

  • outputs are also given warped to the ICBM 2009a Nonlinear Symmetric MNI space (mni_icbm152_t1_tal_nlin_sym_09a.nii)

Further processing

Stats could be further generated with KUL_MS_lesion_stats.sh

Depedencies

A cuda compatible GPU with at least 6GB memory.

Internally KUL_T1T2FLAIRMTR_ratio uses:

These could be installed using KUL_Linux_Installation