lsoustelle/qMT

Description

A collection of command-line Python-based scripts for voxel-wise computation of:

Joint R1f & Macromolecular Proton Fraction (MPF) maps using a Variable Flip Angle (VFA) & a single {MTw/MT₀} protocol [1] (fit_JSPqMT_CLI.py; "Joint Single-Point qMT").
T₁ maps using the Variable Flip Angle method [2] (fit_VFA_CLI.py).
MPF maps using the original Single-Point MPF method [3] (fit_SPqMT_CLI.py).
MTsat maps as described in [4] (fit_SPqMT_CLI.py).

MRI Sequences

Siemens users: two prototype sequences (greMT [VB17/VE11C/VE11E/VE12U/XA20/XA30/XA31] & vibeMT [XA20/XA30/XA31/XA50]) are available on the Siemens C²P Exchange platform (teamplay).
Bruker users: a prototype SPGR sequence is available upon request @LucasSoustelle.

The user interfaces for both constructor match the specific parameters to be passed in the fit_JSPqMT_CLI.py and fit_SPqMT_CLI.py scripts.

Of common note, the fit_JSPqMT_CLI.py script is intended to be used on MTw data prepared with sine-modulated off-resonance pulses [1].

Additional features

For original SP-MPF computation

As in Ref. [5], a synthetic MT reference can be computed from a 2-points VFA protocol (synt_SPGR_CLI.py).
In fit_SPqMT_CLI.py, two presets of constraint qMT parameters are proposed for adult human brain at 3T [3] and adult mouse brain at 7T [6].

Dependencies

The package is intended to have NIfTI files as inputs; please consider conversion package such as dcm2niix or DICOMIFIER.
Specific Python packages: numpy, scipy and nibabel.
- Tested on Python 3.8.2 (Windows 10, Ubuntu 20.04 & 18.04), with numpy-1.19.4, scipy-1.5.4 and nibabel-3.2.0
- For computational efficiency, it is strongly advised to have a numpy package compiled with MKL (native in conda default channels)

Usage

Example for joint Single-Point qMT:

python3 fit_JSPqMT_CLI.py  \
            MT0.nii.gz,MTw.nii.gz \
            VFA.nii.gz \
            MPF_JSPqMT.nii.gz \
            R1f_JSPqMT.nii.gz \
            --MTw_TIMINGS 12.0,2.1,1.0,30.0 \
            --VFA_TIMINGS 1.0,30.0 \
            --VFA_PARX 6,10,25,50,1 \
            --MTw_PARX 560,4000.0,0.0,1,10,1 \
            --B1 B1_map.nii.gz \
            --mask mask.nii.gz \
            --nworkers 20

Example for original Single-Point MPF:

python3 fit_VFA_CLI.py  \
            *FA{4,10,25}*.nii* \
            T1_map.nii.gz \
            --B1 B1_map.nii.gz \
            --mask mask.nii.gz \
            --nworkers 6 \
            --TR 18.0 --FA 4,10,25
          
python3 fit_SPqMT_CLI.py  \
            MT0.nii,MTw.nii \
            T1_map.nii.gz \
            MPF.nii.gz \
            12.0,2.1,0.2,31.0 \
            560,4e3,0,1,10 \
            --RecoTypePreset 1 \
            --B1 B1_map.nii.gz \
            --mask mask.nii.gz \
            --nworkers 6

See the --help option from each script for detailed information.

References

[1] L. Soustelle et al., Quantitative magnetization transfer MRI unbiased by on‐resonance saturation and dipolar order contributions, MRM 2023

[2] L. Chang et al., Linear least-squares method for unbiased estimation of T1 from SPGR signals, MRM 2008;60:496-501

[3] V. Yarnykh, Fast macromolecular proton fraction mapping from a single off-resonance magnetization transfer measurement, MRM 2012;68:166-178

[4] G. Helms et al., High-resolution maps of magnetization transfer with inherent correction for RF inhomogeneity and T1 relaxation obtained from 3D FLASH MRI, MRM 2008;60:1396-1407

[5] V. Yarnykh, Time-efficient, high-resolution, whole brain three-dimensional macromolecular proton fraction mapping, MRM 2016;75:2100-2106

[6] L. Soustelle et al., Determination of optimal parameters for 3D single‐point macromolecular proton fraction mapping at 7T in healthy and demyelinated mouse brain, MRM 2021;85:369-379

lsoustelle / qMT