Introduction to MRtrix
What is MRtrix?
MRtrix is a software package for analyzing diffusion data. One of the notable advantages of MRtrix over tensor-fitting techniques is their method of constrained spherical deconvolution, or CSD; this method deconvolves the diffusion signal in each voxel into a series of overlapping fiber bundles. This reduces the problem of crossing fibers that can be a confound when fitting a tensor.
In addition to a library of commands created by the MRtrix team, the software also has wrappers for commands used with FSL: in particular, the commands
eddy. If you haven’t already, download and install the fMRI software package FSL.
This course is based on the steps outlined in the MRtrix documentation, especially the “DWI Pre-Processing” and “Constrained Spherical Deconvolution” chapters. Several of the steps and explanations are derived from Marlene Tahedl’s excellent BATMAN tutorial, and in many places I use her file notation. I would also like to thank John Plass of the David Brang lab at the University of Michigan for sharing his scripts with me and answering my questions.
Goals of This Course
This course will teach you the basics of diffusion - how it is collected, and how it is analyzed. You will learn how to do fixel-based analyses to quantify the white matter fiber density within each voxel, and how to create tractograms using probabilistic tractography. Finally, you will learn how to create connectomes and how to visualize the amount of fibers that connect distinct brain regions.
- MRtrix Introduction: Overview of Diffusion Imaging
- MRtrix Tutorial #1: Download and Install
- MRtrix Tutorial #2: Downloading the Dataset
- MRtrix Tutorial #3: Looking at the Data
- MRtrix Tutorial #4: Preprocessing
- MRtrix Tutorial #5: Constrained Spherical Deconvolution
- MRtrix Tutorial #6: Creating the Tissue Boundaries
- MRtrix Tutorial #7: Streamlines
- MRtrix Tutorial #8: Creating and Viewing the Connectome
- MRtrix Tutorial #9: Scripting
- MRtrix Tutorial #10: Group-Level Analysis
- MRtrix Tutorial #11: Fixel-Based Analysis