MONAI Label 0.8.5

Added

• SAM2 as base model (Inference only) for all Radiology/Pathology/Endoscopy usecases.

  • Updated 3D Slicer plugin to support ROI prompts for SAM2
    

• OHIF V3 3.9+

  • Refactored Plugin to support latest OHIF V3 3.9+ version
  • Auto Segmentation
  • Point Prompts (Deepgrow/DeepEdit/SAM2) models
  • Class Prompts (Subset of classes as part of Auto Segmentation)
    

• CVAT

  • Refactored Plugin to support latest CVAT version
  • SAM2 model via MONAI Label
    

• Python 3.12 support

Changed

• CVAT Plugin - MONAI Label will serve the inference logic and nuclio function will only at as as wrapper.
• Base Image for docker is ubuntu:22.04 (for saving the space)
• Documentation Fixes

Removed

• Python 3.8 support

0.8.4

What's New:

• VISTA2D Dependencies Added:

  • Added dependencies for VISTA2D, ensuring enhanced compatibility and performance when using this feature.

• Cellprofiler App and Plugin:

  • The Cellprofiler app has been integrated along with a corresponding plugin, enabling new workflows for image analysis.

Enhancements:

• Release Checks and Updates:

  • Updates and improvements made in preparation for the 0.8.4 release.

• Allow Newer Versions of Dependencies:

  • Newer versions of key dependencies are now supported, improving compatibility across a wider range of environments.

• Replace Deprecated IgniteMetric with IgniteMetricHandler:

  • The deprecated IgniteMetric has been replaced with IgniteMetricHandler to keep up with library updates and prevent potential issues.

• MITK Platform Support:

  • MITK has been officially added as a supported platform, expanding the tool's versatility in medical image processing.

Bug Fixes:

• Fix Non-Standard English Double-Colons:

  • Addressed issues with non-standard English double colons, ensuring improved consistency in documentation and messaging.

• Fix Bug with Nuclio Docker Container Directory:

  • Resolved a bug in the Nuclio Docker container that could cause directory-related errors, ensuring smoother execution in Docker environments.

Dependency Updates:

• Bump requests Library:

  • The requests library was updated from 2.31.0 to 2.32.2, addressing security and bug fixes.

• Bump urllib3 Library:

  • The urllib3 library was updated from 2.2.1 to 2.2.2 to fix known issues and maintain security compliance.

• Pre-commit Hooks Update:

  • Updated pre-commit hooks to the latest versions, ensuring consistency in code linting and formatting across the project.

Other Changes:

• Add Test Case for OpenCV Installation:

  • Introduced a test case to ensure both scenarios are enabled when OpenCV is installed, improving test coverage.

• Fix cv2 Dependency Issues:

  • Resolved conflicts related to the cv2 library to ensure smooth functionality in different environments.

0.8.3

What's Changed

• Add no scribbles model
• Mark translatable strings MONAILabel 3D Slicer Extension
• Bump actions/setup-node
• Fixing #1601 - missing server URL error when doing inference
• Updated pydantic version
• Bump github/codeql-action
• Bump actions/upload-artifact
• Bump actions/setup-python
• Update version with remove requests and requests-toolbelt version limit
• Adding save label function to xnat datastore
• Bump actions/cache
• Bump codecov/codecov-action
• Update package dependencies
• Fix response encoding to explicitly use UTF-8 for HTTP response data
• Update fastapi version
• update github monai model zoo config
• Update blossom-ci list
• Downgraded filelock version
• Fixed cv2
• Jwt package for 0.8.2
• Update bundle versions to avoid conflicts
• Update more bundle versions

MONAI Label 0.7.0

Added

• Multi Users authentication and KeyCloak Integration

  • MONAI Label APIs with Keycloak integration for user authentication and role based access.
  • Login support for 3D Slicer via MONAILabel + KeyClock.

• Whole Body CT segmentation

  • Segmenting 104 anatomies in 4 seconds!

• MONAI Bundle Support Improvements

  • Support visualization of bundle config option.
  • Enhancement of monai-zoo access.
  • Support bundle downloading from NGC.
  • Enhacement of multi-gpu training of bundles.

• New MONAI Label Tutorial Series

  • Quickstart tutorials and installation instructions in notebooks.

• Documentation enhancements

Changed

• Updated pretrained models:

  • segmentation
  • deepedit

• CI/CD and tests

  • blossom CI/CD and pre-merge pipeline enabled.
  • Increased Unit tests coverage to 80%.

Removed

MONAI Label 0.6.0

Added

• Pathology Models
  • NuClick annotation model
  • Nuclei classification model
  • HoverNet model
• QuPath Extension: 0.3.1 | Demo
  • User experience enhancements
  • MONAI Label specific Toolbar actions
  • Drag and Drop ROI to run auto-segmentation models
  • Single click to run interaction models (NuClick)
  • Support Next Sample/ROI for Active Learning
• Experiment Management
• 3D Slicer: Detection model support in MONAI Bundle App for Radiology use-case
• Multi-GPU/Multi-Threaded support for Batch Inference

Changed

• Support latest version of bundles in MONAI Bundle App
• Upgrade vertebra pipeline
• MONAI version >= 1.1.0

Removed

• DeepEdit model for Nuclei segmentation in Pathology

Pretrained Model Weights

THIS IS NOT A RELEASE PAGE

Download Pretrained model weights (TEST/DEMO purpose only) for the sample apps from here

MONAI Label 0.5.2

Added

• Bundle support (Endoscopy Sample App)
  • Tool Tracking segmentation model
  • InBody vs OutBody classification model
• Rest APIs to download latest trained model and stats
• Interactive NuClick segmentation in DSA | Demo
• Support for MONAI 1.0.1

Changed

• Option to disable DICOM to NIFTI conversion for OHIF usecase
• Proxy URL fixes for GCP + DICOMWeb support
• 3D Slicer UI Improvements
  • Optimize window size for options/config
  • Fix download label feature to fetch original labels
• Fixes for multi label output of segmentation_nuclei model (Pathology)
• Improvements on MONAI Bundle App
  • support local bundles (pre-downloaded)
  • support customized scripts

Removed

• Remove option to run ALL Training Tasks in 3D Slicer (Deprecated)

MONAI Label 0.5.1

Added

• Endoscopy Sample App
  • Tool Tracking segmentation model | Demo
  • InBody vs OutBody classification model | Demo
  • DeepEdit interaction model
  • CVAT Integration to support automated workflow to run Active Learning Iterations
• Multi Stage vertebra segmentation in Radiology App

Changed

• Improving performance for Radiology App
  • Support cache for pre-transforms in case repeated inference for interaction models
  • Support cache for DICOM Web API responses
  • Optimize pre transforms to run GPU for max throughput
  • Fix DICOM Proxy for wado/qido
• Improvements for Epistemic (v2) active learning strategy
• Support for MONAI 1.0.0 and above
• Fixes Scribbles to support MetaTensor

Removed

• TTA Based active learning strategy is deprecated

MONAI Label 0.4.2

Added

• MONAI Bundle App - Pull compatible bundles from MONAI Zoo
  • spleen_ct_segmentation
  • spleen_deepedit_annotation
  • others

Changed

• Support for MONAI 0.9.1

MONAI Label 0.4.1

Changed

• Fix MONAI dependency version to 0.9.0