What's New In XNAT

XNAT 1.10.2 extends what XNAT can do with DICOM data, and arrives alongside Container Service 3.9.0 and recent releases of the OpenID Authentication Plugin and the XNAT Desktop Client.

Snapshots for more kinds of scans

Snapshots and thumbnails now render for JPEG-compressed scans, segmentations, whole-slide imaging and other series that mix single- and multi-frame instances, and scans whose objects all went to the secondary DICOM catalog.

A new pixel redaction engine

alterPixels now redacts without its memory use growing with the image, keeps the transfer syntax of uncompressed and lossless objects where it can, and uses the fill value the script asks for. Anonymization streams pixel data from disk, so objects with Pixel Data over 2 GiB can be anonymized and redacted.

More of the DICOM header

Routing rules, DICOM mappings and the DICOM dump can now use elements that come after Pixel Data, including private tags in groups 0x8000 and above.

Container Service 3.9.0

Container Service can now clean up the build directories that finished containers leave behind, on a daily schedule or on demand.

Also recently released

• OpenID Authentication Plugin 1.6.0: token claim checks, bearer-token authentication for REST requests, account linking and auto-login.

• XNAT Desktop Client 3.2.6: a native build for Apple Silicon Macs, signed installers, and per-modality control of image anonymization.

Full details are in the XNAT 1.10.2 Release Notes.

XNAT 1.10 represents a major technology update under the hood, offering support for Java 21 as well as a major update to the dcm4che processing library used in importing and anonymizing DICOM image session data. Read the XNAT 1.10.0 Release Notes for more details.

This 1.10.2 update builds on changes from XNAT 1.10.0 and 1.10.1 which offered new functionality, including (but not limited to) the following: 

Dynamic Data Types can be created in XNAT 1.10 without a Tomcat Restart

In a major new feature addition, XNAT now supports the ability to create Dynamic Data Types. Used in combination with Custom Forms, site administrators now have the flexibility to support whole new classes of experiments and image assessors without requiring backend development or a server restart.

Improved Support for Load Balancing with the Distributed Events Plugin

For load-balanced XNAT environments where multiple front end servers are connected to a single PostgreSQL database, the Distributed Events Plugin makes for a more seamless user experience. This plugin helps propagate events related to users and data across nodes, providing greater accuracy and fidelity regardless of which node a user is logged into.

Updated & Modernized Dependency Libraries in XNAT 1.9

XNAT 1.9.0 represents a major effort to update foundational libraries and dependencies on which XNAT is built. The motivation for this effort is to move toward compliance with a SOC2 security audit. Read the XNAT 1.9.0 Release Notes for more details. 

These updates also prescribe a set of updates to core XNAT plugins for compatibility with these updates, including: 

• Container Service 3.6.0

• Batch Launch Plugin 0.7.0

• XSync 1.8.0

• DICOM Query Retrieve 2.1.0

Independent plugin developers should review our Guide to Updating Plugins for Changes in XNAT Dependencies to determine whether changes will be required for your plugins. 

Deprecation of the Pipeline Engine 

Earlier releases made XNAT more independent of the Pipeline Engine by removing the dependency on the AutoRun Pipeline by introducing the Scan Snapshot Generation API. As of XNAT 1.9, the Pipeline Engine is only available as a plugin. This move has been made to remove dependencies on libraries that are only in use in the Pipeline Engine, and that can contain 

See: XNAT Pipeline Engine Plugin

Deprecation of ML & Datasets Plugins

The XNAT ML suite of plugins is no longer being supported, and we are now offering mechanisms to safely remove their functionality from your XNAT installations, should you wish to do so. Read the full deprecation notice in the XNAT ML space for more details.

In the meantime, XNAT community development into the ML world has taken different forms:

• The XNAT-OHIF Viewer plugin, developed at ICR UK, introduced support for NVIDIA’s MONAI Label suite of AI-Assisted Annotation tools

• The XNAT JupyterHub Integration plugin, developed at Washington University in St Louis, offers expanding support for mounting XNAT data into a Jupyter Notebook and performing exploratory model development

Support for Microscopy in the XNAT OHIF Viewer

As of version 3.7.0 of the XNAT-OHIF Viewer, DICOM Whole Slide Images (WSI) including microscopy data are now supported in the viewing panel.

Read the full news announcement here: Container Service 3.5.0 and XNAT-OHIF 3.7.0 Releases

Related: XNAT OHIF Viewer

Support for Docker Engine 25+ included in Container Service 3.5.0 and up

Version 3.5.0 of the Container Service addresses a critical need, offering support for running on Docker Engine v25+. This version of the Docker Engine deprecates several APIs and features that required some re-engineering of the Container Service backend. However, other than updating to the latest plugin version, there is no other update required by XNAT admins to their containers. 

Read the full news announcement here: Container Service 3.5.0 and XNAT-OHIF 3.7.0 Releases. 

Related: Container Service

Support for Multiple Open Dashboards and Notebooks added to JupyterHub Plugin

JupyterHub 1.2.0 adds support for running multiple notebooks and/or dashboards simultaneously, as well as other improvements. See the XNAT JupyterHub 1.2.0 Release Notes for more details. 

Related: 

Full Archive of XNAT Release Notes