Diffusion Imaging In Python - Documentation

DIPY is the paragon 3D/4D+ imaging library in Python. It contains generic methods for spatial normalization, signal processing, machine learning, statistical analysis and visualization of medical images. Additionally, it contains specialized methods for computational anatomy including diffusion, perfusion and structural imaging.

DIPY is part of the NiPy ecosystem.

Quick links

Get started

New to DIPY? Start with our installation guide and DIPY key concepts.

Installation

Tutorials

Browse our tutorials gallery.

Examples

Recipes

How do I do X in DIPY? This dedicated section will provide you with a quick and direct answer.

Recipes

Workflows

Not comfortable with coding? We have command line interfaces for you. An easy way to use DIPY via a terminal.

DIPY Workflows Interfaces

Theory

Back to the basics. Learn the theory behind the methods implemented in DIPY.

Theory and concepts

Developer Guide

Saw a typo? Found a bug? Want to improve a function? Learn how to contribute to DIPY!

DIPY Developer Guide

API reference

A detailed description of the DIPY public Python API.

API Reference

Workflows API reference

A detailed description of all the DIPY workflows command line.

Command Line Utilities Reference

Release notes

Upgrading from a previous version? See what’s new and changed between each release of DIPY.

A quick overview of features

Get help

Need help with your processing? Ask us and a large neuroimaging community.

https://github.com/dipy/dipy/discussions

Highlights

DIPY 1.12.0 is now available. New features include:

• NF: FORCE reconstruction model.

• NF: New BiasField correction method.

• NF: Parallel EuDX tractography.

• NF: Intermediate map for symmetric diffeomorphic registration.

• NF: StatefulSurface class to handle surfaces.

• NF: Multiple new workflows (dipy_fit_msmtcsd, dipy_brain_mask, dipy_cluster_streamlines, dipy_fit_powermap, dipy_fit_fwdti).

• ENH: Cythonized AK, RK, and KFA kurtosis computations.

• RF: TRX as default file format for tractography outputs.

• RF: Adoption of pathlib across workflows.

• RF: Refactoring of nlmeans denoising (classic and blockwise variants).

• Added support for Python 3.14.

• Drop support for Python 3.10.

• Documentation update.

• Closed 324 issues and merged 173 pull requests.

See Older Highlights.

Announcements

• DIPY 1.12.0 released March 14, 2026.

• DIPY 1.11.0 released March 15, 2025.

• DIPY 1.10.0 released December 12, 2024.

See some of our Past Announcements