16.3.x: New DICOM Viewer Features, Improved Job Management and Performance Enhancements
For many of you, much has changed since your last upgrade. Version 16.3.5 brings a host of new functionality and performance improvements. This latest update includes:
- DICOM Viewer Updates
- Basic Digital Pathology Support
- MHD / Raw Support
- Undo / Redo Annotation Actions
- Speed and Performance Improvements
For those that are upgrading from version 14, don't miss these feature announcements:
- Version 16.0.0
- Gear Rule Enhancements
- View and Manage Analyses at all Levels, Including Acquisition-Level
- Version 15.6.0
- File Versioning
- New DICOM Viewer Features and Improvements
- User Interface Enhancements
New Jobs Log
Version 16.3 includes a new jobs log accessible to all users in the "Compute" section of the lefthand navigation bar. This much-requested feature gives users the ability to view and manage their jobs from the Flywheel UI.
The new jobs log features a more robust list with additional job information, infinite scroll, and column level filtering for easy navigation.
DICOM Viewer Updates
We're continuing to improve and enhance our native DICOM viewer with support for additional data types, performance upgrades, and an improved user experience.
Digital Pathology
The viewer now gives you the ability to load and tile through digital pathology DICOM images. This first phase of support also provides basic annotation tools.
Whole slide imaging data was converted to DICOM using the WSI to DICOM Gear, which supports a number of commonly used formats.
MHD / Raw Support
The viewer now supports the viewing and basic annotation of .mhd files. These images are displayed in a three-viewport configuration showing the axial, coronal, and sagittal planes.
Undo / Redo
This latest update provides a significantly improved experience for users tasked with annotating and labeling images. Actions can easily be undone and redone (up to 10 actions) by clicking on the corresponding icon in the measurement panel or through hotkeys: "ctrl + z" and "ctrl + shift + z".
For detailed version-by-version updates including bug fixes, see Release Notes.
16.0.0: Gear Rule Enhancements
Gear Rules just got a lot more powerful with this latest update that allows rules to match on:
- Filename
- File type
- File tag
- Analysis label
- Gear name
- Gear version
Additionally, gear rules now allow you to specify at which level of the Flywheel container to look for files, including analysis outputs. These enhancements will allow you to automate multi-step pipelines and workflows.
As part of this improvement, we've also made it easier to view and manage analyses in the UI. When viewing analyses at any level, you now have the option to also view analyses associated with its children containers.
Acquisition level analyses can now be viewed in the Flywheel UI!
For more information see our Gear Rules article.
15.6.0: File Versioning, DICOM Viewer Updates, UI Enhancements, and Performance Improvements
Version 15.6.0 includes a host of new functionality and performance improvements. This update includes:
- File Versioning
- New DICOM Viewer Features and Improvements
- User Interface Enhancements
- CLI Updates
- Fixes
File Versioning
Flywheel now offers even greater support for reproducibility and regulatory compliance by introducing numbered versioning for all file data. Now, file versions will be automatically recorded in all provenance, analyses, and jobs logs. This feature also gives you better control of your data by allowing you to view, restore and delete previous versions of files.
See this article for the full details.
DICOM Viewer Features and Improvements
Updating and improving our integrated viewer is a key focus for Flywheel in 2021. Version 15.6.0 brings some exciting new functionality as well as some key improvements to performance and consistency.
Overlay NIfTI Images
The ability to superimpose different images in the viewer is an important requirement to many of your workflows. We're happy to announce that we've extended the viewer's capabilities to support the application and management of multiple overlays directly in the viewer. Within the viewer, you can now navigate the Flywheel hierarchy to select segmentation overlays and masks.
For example, after loading a NIfTI image in the viewer, you can easily navigate to its associated FreeSurfer analysis object and load the output overlays.
Additional Enhancements
Version 15.6.0 also includes significant improvements to viewer performance, consistency, and image-type support. Specifically:
- Updates to caching and indexing logic for improved loading speed
- More robust handling of datasets missing expected tag values
- Support for encapsulated PDF files
- Support for displaying Siemens mosaic images
User Interface Enhancements
We're continuing to update the UI to improve usability and overall user experience. Version 15.6.0 includes additional sorting and filtering options as well as improvements in how you can view and interact with your data in the UI.
Sorting and Filtering
As many of you have told us, it can be challenging to locate the acquisitions you need when working with data-heavy sessions. Version 15.6.0 adds additional sorting options for acquisitions in addition to the existing sorting and filtering options for acquisition files.
We've also added sorting and filtering to all attachment interfaces to make it easy to locate the files you need.
Viewing Timezone Data
As several users have reported, attempting to locate scans based on timestamps in the UI is challenging when those timestamps are converted to local time by the browser. To alleviate this confusion, we've added a toggle that allows you to view DICOM timestamps with their corresponding timezones at the time and place of capture.
Data Views Enhancements
Since its release last year, many of you have adopted Data Views as an important component of your workflows. With that, we're committed to continuing to improve and extend its functionality to meet your needs. Version 15.6.0 includes:
- The ability to adjust column widths in the UI
- Cell value displayed on mouse hover
- Backend updates to improve performance on creating, launching and exporting larger Data Views
CLI Updates
Test De-id Locally
Because proper de-identification is absolutely critical to your workflows, we've made it easier to create and test your de-id profiles before your data reaches Flywheel. We've introduced a top-level command that allows you to generate a sample template that you can use as a starting point and a simplified method for testing your template. Simply point to a sample dataset, the profile you'd like to test, and the folder where you'd like the output report to be saved.
| fw deid ~/DICOMS ~/Profile/deid_profile.yaml ~/TestOutput |
See this article for the full details