Skip to content

Latest commit

 

History

History
65 lines (43 loc) · 4.67 KB

README.md

File metadata and controls

65 lines (43 loc) · 4.67 KB
Raw

dMRI & Chronic Pain

The starting point and aim of this project is to use diffusion MRI to improve our understanding of chronic pain and its treatment.

Dataset

Raw DWI images are acquired with the same protocol to offer a homogeneous dataset: b-values of 300, 1000 and 2000 with a resolution of $2\times 2\times 2 mm^3$.

Pipelines and scripts

Before running the pipelines and scripts it is required to install Nextflow, Singularity, Scilpy, and Ants. A general installation procedure can be found here and more specific instructions, for each pipeline, can be found in their respective sections.

If the pipelines are ran on Linux, it is recommended to use Singularity. Pre-built singularity images and container related files for SCILUS flows can be found on the github repository containers-scilus or on the SCIL website.

dMRI QC

Using the dmri_qc nextflow pipeline it is possible to check DWIs and the processed DWIs. A more in depth description and installation intruction can be found here.

Resources
Example command

nextflow run dmriqc-flow-0.1.2/main.nf -profile input_qc --root input/ -with-singularity singularity_dmriqc_0.1.2.img -resume --raw_dwi_nb_threads 10

TractoFlow

Using the TractFlow nextflow pipeline it is possible to compute necessary derivatives: DTI metrics, fODF metrics. The script used to run is run_tractoflow

Resources
Example command

nextflow run main.nf --input <DATASET_ROOT_FOLDER> --dti_shells "0 300 1000" --fodf_shells "0 1000 1200" -with-singularity

Outputs

DTI metrics: The Diffusion Tensor Imaging metrics computed are: the axial diffusivity (AD), fractional anisotropy (FA), geodesic anisotropy (GA), mean diffusivity (MD), radial diffusivity (RD), tensor, tensor norm, tensor eigenvalues, tensor eigenvectors, tensor mode, color-FA. Use flag -dti_shell to specify the desired shells t compute DTI metrics. Usually it is recommended to compute DTI metrics using b-values under $b = 1200 mm^2/s$.

fODF metrics: The fiber Orientation Distribution Function metrics computed are: the total and maximum Apparent Fiber Density (AFD), the Number of Fiber Orientation (NuFO) and principal fODFs orientations (up to 5 per voxel). Use flag –fodf_shells to specify the desired shells to compute fODF metrics and flag --sh_order to specify the spherical harmonic order (default is 8). Usually it is recommended to compute fODF metrics using b-values above $b = 700 mm^2/s$.

RecobundleX

Using the rbx_flow nextflow pipeline it is possible to extract white matter fiber bundles of interest. The script used to run is run_rbxflow.

Resources
Example command

nextflow run main.nf -resume -with-singularity scilus-1.2.0_rbxflow-1.1.0.img --input input/ --atlas_config code/rbx-atlas/config.json --atlas_anat code/rbx-atlas/mni_masked.nii.gz --atlas_directory code/rbx-atlas/atlas/