ivim documentation

ivim is an end-to-end python package for intravoxel incoherent motion (IVIM) analysis of diffusion MRI data. It is designed in a modular fashion where all functions involving processing of data has nifti files as input/output. This allows for setting up pipelines where functionality from well-known diffusion MRI packages like FSL and MRTrix3 (Gibbs removal, motion correction, etc) can be combined with IVIM specific functionality available in the current package. The file formats for diffusion MR images are those of FSL (.nii.gz, .bval, .bvec), with extensions in line with this where needed (e.g. .cval for c-values describing velocity encoding).

Here you find the documentation of release 1.0, with last update Dec 03, 2025.

Note

The package consistently use units mm, s and T. This is in line with the typical units for diffusivity (mm²/s) and b-value (s/mm²). Extending to velocities and c-value this gives suitable units of mm/s and s/mm, respectively. It does imply that sequence timing must be given in s rather than ms, and most importantly that gradient strengths are given in T/mm rather than mT/m. The typical user would however rarely deal with the functions involving gradients strengths are input/output.

Getting started

• Installation
• Examples

Detailed information

• Models
• Pulse sequences
• Protocol optimization
• Simulation
• Input/Output
• Preprocessing
• Model fitting
• Additional examples