NA-MIC Project Weeks

Back to Projects List

Image visualization in Google Colab

Key Investigators

• Andrey Fedorov (BWH, USA)
• Andras Lasso (Queen's, Canada)
• Justin Kirby (Frederick National Lab, USA)

Project Description

In this project we will experiment with different approaches to implementing 3d visualization of radiology images and overlaying segmentations in Google Colab. The primary motivation for this project is to support visualization of images from Imaging Data Commons along with user-generated analysis results in Google Colab, which has zero prereqisites for installing software on user’s computer.

Objective

1. Develop a publicly available Colab Notebook to demonstrate how to use trame-slicer for visualizing images and segmentations.
2. Use examples from Imaging Data Commons for demonstration.
3. Document and report any issues identified.

Approach and Plan

1. Read up on what trame-slicer is and how to use it.
2. Start from the instructions put together by Andras in https://github.com/KitwareMedical/trame-slicer/issues/81.

Progress and Next Steps

1. Prepared notebooks demonstrating the use of trame-slicer, ipyniiview and Steve’s desktopia: https://github.com/ImagingDataCommons/IDC-Tutorials/pull/104
   • ipyniiview appears to be most robust, lightweight and fast viewer to use in Colab notebook (not tested outside of Google)
   • trame-slicer developers identified directions to improve performance and simplify usage, will monitor
2. Contributed new tutorials to IDC-Tutorials repo, updated “getting started” with using ipyniiview: https://github.com/ImagingDataCommons/IDC-Tutorials/pull/104
   • updated intro tutorial with visualization

Illustrations

Result of running this IDC tutorial notebook

No response

Background and References

• https://colab.research.google.com/

No response