MRtrix 2016
Introduction
MRtrix3 provides a set of tools to perform various types of diffusion MRI analyses, from various forms of tractography through to next-generation group-level analyses. It is designed with consistency, performance, and stability in mind, and is freely available under an open-source license. 1
Installed version(s)
The following versions are installed and currently available...
... on environment hpc-env/8.3:
- MRtrix/3.0.2-intel-2019b-Python-3.7.4
- MRtrix/3.0.3-intel-2019b-Python-3.7.4
Loading MRtrix
To load the desired version of the module, use the module load command, e.g.
$ module load hpc-env/8.3 $ module load MRtrix
Always remember: these commands are case-sensitive!
Using MRtrix
MRtrix comes with a lot of binaries which you can make use of. To getyou a quick overview about the binary content, you can take a look at the softwares bin folder aftert loading the module:
$ ml MRtrix $ ls -la $EBROOTMRTRIX/bin
Most of the displayed executables do come with a help function if you just call the corresponding binary without any arguments entailed. For example:
$ mrtransform
MRtrix 3.0.3 mrtransform Oct 24 2022
mrtransform: part of the MRtrix3 package
SYNOPSIS
Apply spatial transformations to an image
USAGE
mrtransform [ options ] input output
input input image to be transformed.
output the output image.
DESCRIPTION
If a linear transform is applied without a template image the command will
modify the image header transform matrix
FOD reorientation (with apodised point spread functions) can be performed
if the number of volumes in the 4th dimension equals the number of
coefficients in an antipodally symmetric spherical harmonic series (e.g.
6, 15, 28 etc). For such data, the -reorient_fod yes/no option must be
used to specify if reorientation is required.
The output image intensity can be modulated using the (local or global)
volume change if a linear or nonlinear transformation is applied. 'FOD'
modulation preserves the apparent fibre density across the fibre bundle
MRtrix 3.0.3 mrtransform Oct 24 2022
mrtransform: part of the MRtrix3 package
SYNOPSIS
Apply spatial transformations to an image
USAGE
mrtransform [ options ] input output
input input image to be transformed.
output the output image.
DESCRIPTION
If a linear transform is applied without a template image the command will
modify the image header transform matrix
FOD reorientation (with apodised point spread functions) can be performed
if the number of volumes in the 4th dimension equals the number of
coefficients in an antipodally symmetric spherical harmonic series (e.g.
6, 15, 28 etc). For such data, the -reorient_fod yes/no option must be
used to specify if reorientation is required.
The output image intensity can be modulated using the (local or global)
volume change if a linear or nonlinear transformation is applied. 'FOD'
modulation preserves the apparent fibre density across the fibre bundle
width and can only be applied if FOD reorientation is used. Alternatively,
non-directional scaling by the Jacobian determinant can be applied to any
image type.
If a DW scheme is contained in the header (or specified separately), and
the number of directions matches the number of volumes in the images, any
transformation applied using the -linear option will also be applied to
the directions.
When the -template option is used to specify the target image grid, the
image provided via this option will not influence the axis data strides of
the output image; these are determined based on the input image, or the
input to the -strides option.
Affine transformation options
-linear transform
specify a linear transform to apply, in the form of a 3x4 or 4x4 ascii
file. Note the standard 'reverse' convention is used, where the transform
maps points in the template image to the moving image. Note that the
reverse convention is still assumed even if no -template image is supplied
-flip axes
flip the specified axes, provided as a comma-separated list of indices
(0:x, 1:y, 2:z).
-inverse
apply the inverse transformation
Documentation
The full documentation can be found here.
