QSM Tools

List of open-source quantitative susceptibility mapping (QSM) MRI software packages and data.

Please open an issue or pull request if you believe any corrections or additions to the list should be made.

• Tools
• Data

Tools

• SEPIA
  Details

  • Language: MATLAB
  • License: MIT
  • Description:
    • Provides a GUI with other tools to create pipelines.
    • Tools:
    • MEDI toolbox
    • STI suite
    • FANSI toolbox
    • SEGUE
    • MRI susceptibility calculation methods
    • ROMEO

• MEDI Toolbox
  Details

  • Language: MATLAB
  • License: Academic-only, signup required
  • Methods:
    • Morphology Enabled Dipole Inversion (MEDI)
    • Truncated singular value decomposition (TSVD)
    • Truncated k-space division (TKD)
    • Iterative SWIM (iSWIM)
    • TV with Split-Bregmann (TVSB)

• Hongfu Sun's QSM software
  Details

  • Language: MATLAB
  • License: MIT
  • Methods:
    • Total variation dipole inversion

• Berkin Bilgic's QSM software collection
  Details

  • Language: MATLAB
  • License: Unknown
  • Methods:
    • Single-Step QSM with Total Generalized Variation regularization
    • Fast Total Generalized Variation regularized QSM
    • Reconstruction for Multi-orientation QSM and Susceptibility Tensor Imaging
    • Single-Step QSM with l2-Regularization
    • Fast l1-Regularized QSM with Magnitude Weighting and SHARP background filtering
    • Closed-form l2-Regularized QSM
    • l1- and l2-Regularized QSM and PDF background filtering

• TGV QSM Algorithm
  Details

  • Language: Python
  • License: Unknown
  • Methods:
    • Total generalized variation

• QSM Pipeline
  Details

  • Language: Docker and Singularity image bundles.
    • Core QSM tools written in Python
  • License: Unkown
  • Methods:
    • Total generalized variation

• Susceptibility Imaging Toolkit
  Details

  • Language: Python
  • License: Unknown
  • Methods:
    • QSM inversion

• STI Suite(alternate link)
  Details

  • Language: MATLAB
  • License: Non-commercial academic use only, email signup required
  • Methods:
    • Quantitative Susceptibility Mapping without spatial constraints

• Quantitative Susceptibility Mapping for MRI part 1
  Details

  • Language: MATLAB
  • License: MIT
  • Methods:
    • Closed-form l2-Regularized QSM (Bilgic et al.)

• Quantitative Susceptibility Mapping for MRI part 2
  Details

  • Language: MATLAB
  • License: MIT
  • Methods:
    • L1-regularized magnitude weighted QSM (Bilgic et al.)

• QSM SSTV
  Details

  • Language: MATLAB
  • License: Unknown
  • Methods:
    • Single step total variation

• pyQSM
  Details

  • Language: Python
  • License: Unknown
  • Methods:
    • Dipole inversion

• QSM L0 reconstruction algorithm
  Details

  • Language: MATLAB
  • License: Unkown
  • Methods:
    • L0 reconstruction

Data

• QSM reconstruction challenge 2016
  Details

  • Field strength: 3.0 Tesla
  • Resolution: 1.06 mm
  • Subject: Healthy volunteer
  • Other:
    • A reference susceptibility map computed using the susceptibility tensor imaging algorithm on data acquired at 12 head orientations.

• STI Suite
  Details

  • Human Brain Single Echo 2D
    • Field strength: 3.0 Tesla
    • Resolution: 1.8 x 1.8 x 3.0 mm³
    • Subject: Human
  • Human Brain Single Echo 3D
    • Field strength: 3.0 Tesla
    • Resolution: 1.25 x 1.25 x 2.0 mm³
    • Subject: Human
  • Human Brain Multi Echo 3D
    • Field strength: 3.0 Tesla
    • Resolution: 1.25 x 1.25 x 2.0 mm³
    • Subject: Human
  • Mouse Brain 3D
    • Field strength: 9.4 Tesla
    • Resolution: 1.0 x 1.0 x 2.0 um³
    • Subject: Human

• MEDI Toolbox
  Details

  • Field strength: Unkown
  • Resolution: 0.9375 x 0.9375 x 2.0 mm³
  • Subject: Human and simulation
  • Other:
    • Siemens and GE datasets

• Quantitative Susceptibility Mapping (QSM): Decoding MRI Data for a Tissue Magnetic Biomarker
  Details

  • Field strength: Unkown
  • Resolution: 0.9375 x 0.93785 x 1.0 mm³
  • Subject: Human and simulation

• Susceptibility Imaging Toolkit
  Details

  • Field strength: Unknown
  • Resolution: Unknown
  • Subject: Human