Diffusion Imaging in Python

Diffusion Imaging in Python, known as DIPY, is an open-source Python library for 3D and 4D medical imaging, with a particular focus on diffusion MRI of the brain. It provides general methods for spatial normalisation, signal processing, machine learning, statistics and visualisation, plus specialised tools for computational anatomy.

What Diffusion Imaging in Python offers

• Diffusion models — single-shell methods such as DTI, CSA and CSD, and multi-shell methods such as DKI, SHORE and MAPMRI.
• Registration — affine, diffeomorphic 2D/3D and streamline-based registration.
• Tractography — probabilistic, deterministic and PFT tracking of white matter pathways.
• Denoising and preprocessing — Patch2Self, non-local means, Gibbs unringing, brain extraction and SNR estimation.
• Visualisation — ODF display and interactive tractogram viewing.
• Command-line workflows — every algorithm is available through a CLI, and users can add their own.

Who it works with

The library serves neuroimaging researchers and developers; partners offer consulting, custom development and training for organisations that need help.

Key facts

• Type: open-source software
• Language: Python
• Publications: methods including BUAN, EVAC+ and BundleWarp have appeared in peer-reviewed journals