Working with Processing Pipelines

Quick Index

Pipelines in XNAT are powerful mini-applications that can be run on your project data, to aid in complex processing or leverage the power of large computing clusters. Some pipeline-enabled workflows are carried out automatically without any human intervention, like auto-QC'ing images as they are added to your project archive. Others require a person to do manual steps, such as drawing a region of interest.  

When pipelines are integrated into your project, you can:

  • set up project-based workflows with project specific and experiment specific parameters,
  • track a pipeline and send email notifications, and
  • capture provenance information as the pipeline executes. 
Only XNAT Site Administrators can install new pipelines, but once installed on your site, any project owner can add them to their project. Then, any user with project permissions can run that pipeline on your data. 

Adding Pipelines to your Project

As the project owner, you can add any pipeline that is active in XNAT. This is done from the project report page, under the PIPELINES tab. Here is a step-by-step walkthrough.

  1. From within a project, click on the Pipelines tab
  2. Click Add More Pipelines
  3. Click Add next to the pipeline to be added - in this case, we'll use the Web Based QC Image Creator that ships standard with XNAT. If there is no pop-up when you click add, the problem may be that the pipeline was added with backslashes '\' in the path. Try re-adding the pipeline using forward slashes '/' in the path instead.
  4. Insert any needed parameters into the pipeline configuration form. 
  5. Click Submit. A window will come up with the message: "The pipelines for the project were successfully modified"

You can then close the popup window, and your pipeline will be active on your project. 

Running Pipelines on your Project

For the most part, pipelines do not directly modify your experiment data; instead, they add additional data to your project as Assessors. (Related: Understanding the XNAT Data Model)

To run a pipeline, you must first have installed a pipeline, of course. Then, navigate to the experiment report page (for example, an MR Session page) in XNAT.  

  • Click Build within the Actions box at the right. 
  • Select the pipelines to run from the pop-up window

  • Click Submit
  • A new window will open with the message

Your pipeline is now running!