[FEATURE] Design dMRI preprocessing and analysis

[tclose]

Feature

Design Arcana Analysis class that implements diffusion MRI preprocessing and analysis up to biomarkers.

User requested

QUT

• Motion realignment
• Eddy current/artefact correction
• Coregistration of subject T1/T2 to diffusion space
• The type of diffusion processing would be protocol dependent but some basic modifiable pipelines can be offered (tensor for old school data, csd, etc)

UniMelb

• MRTrix
• FSL -a) TOPUP, b) Eddy
• QSIprep

Swinburne

• Diffusion MRI preprocessing
  Input: T1, 2 or 3-shell dwi
  Output: Three 3D volumes (with FODs for WM, GM and CSF), brain mask based on dwi
  Suggested features: option to process HCP "minimally preprocessed" data (see first steps in: https://github.com/SwinburneNeuroimaging/HCP-dMRI-connectome)
  Recommended tools: mrTrix, QSIprep BIDS app
• Diffusion MRI QC
  Input: 2 or 3-shell dwi
  Output: QC metrics
  Recommended tools: eddyqc (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/eddyqc), QSIprep, ExploreDTI (MATLAB)
• NODDI (neurite orientation dispersion and density imaging)
  Input: 2 or 3-shell dwi
  Output: 3D volume with NODDI measures, brain mask based on dwi
  Suggested features: option to process HCP "minimally preprocessed" data
  Recommended tools: QSIPrep BIDS app, DIPY
• Voxelwise diffusion MRI measures (FA, RD, AD)
  Input: output of Diffusion MRI preprocessing or 2 or 3-shell dwi
  Output: FA, RD and AD 3D volumes
  Suggested features: option to process HCP "minimally preprocessed" data
  Recommended tools: mrTrix, FSL, QSIprep BIDS app

Benefit Hypothesis

If comprehensive dMRI Analysis classes are implemented, then they will provide a convenient method for users of AIS/NIF to analyse diffusion MRI data, which is

• based on state-of-the-art methods
• reproducible (i.e. full provenance, portable to other datasets/systems)
• scales efficiently to large datasets

Acceptance Criteria

• All of the requested features that have been approved by AIS AOC are implemented by the Analysis class
• Tests cover all of the implemented pipelines
• Each pipeline lists all relevant citations, specifies resource requirements and has descriptive doc strings
• Selected workflows are deployed in AIS pipelines for XNAT CS

Tasks

Preview Give feedback

No tasks being tracked yet.