nD Image Visualization with Open-Source Tools

This repository contains a few demo notebooks for creating interactive 3D visualizations using Python.

• ↗️ PyVista Example

• ↗️ Itkwidgets Example

• ↗️ Dash VTK Example

Setup & Installation

Install the Python requirements:

pip install -r requirements.txt

Start jupyter lab from the terminal:

jupyter lab

List of 3D image visualization tools

In addition to Jupyter notebook tools, many other 3D rendering software packages exist. Here are a few that we recommend (in alphabetical order):

• Dash VTK
• Fiji - 3D viewer
• Fiji - Volume viewer
• Itkwidgets
• K3D
• Napari
• Neuroglancer
• Paraview
• PyVista
• stackview
• tif2blender
• vedo
• Viv
• Vizarr

Which tool for which use case?

Features

Software	3D+time	Multichannel	Large data	Volume rendering	Projections (mip)	Isosurface / meshes	Glyphs	Intuitive	Scriptable
Fiji Volume Viewer	-	-	-	✅	✅	-	-	✅	-
Fiji 3D Viewer	✅	✅	-	✅	-	✅	-	✅	-
Napari	✅	✅	✔️	✔️	✅	✅	✅	✅	✅
PyVista	-	✔️	✔️	✅	-	✅	✅	✔️	✅
Neuroglancer	✅	✅	✅	✔️	✔️	✅	✔️	-	✔️
Paraview	✅	✔️	✔️	✅	-	✅	✅	-	✔️

✅ Yes ✔️ Sort of

Pros and cons

Fiji Volume Viewer

• ✅ Ideal for Fiji users
• ✅ Good control over the 3D rendering
• 🔴 No glyphs or overlays (masks, points, vectors...)
• 🔴 4D (3D+time or multichannel) not supported (?)
• 🔴 Not controllable programmatically

Napari

• ✅ Ideal for nD: 3D+time, multichannel
• ✅ Ideal for overlays: masks, points, vectors...
• ✅ Controllable programmatically
• 🔴 No fine control over the transfer function

PyVista

• ✅ Ideal for reproducible visualizations in Python
• ✅ Good control over the 3D rendering
• ✅ Desktop or web-based
• 🔴 Not as interactive as other tools

Neuroglancer

• ✅ Ideal for Zarr and large images
• ✅ Visualizations can be shared simply with a URL
• 🔴 Not as intuitive as other tools