Skip to content

Release Notes

0.1.1 [2025-10-24]

Enhancements:

  • Added support for stochastic transformations with multiple iterations via new number-of-iterations config parameter
  • Added support for Compose transforms starting with LoadImage (in addition to LoadImaged)

Fixes:

  • Fixed output directory configuration for SaveImage and SaveImaged transforms to use folder_layout.output_dir

Maintenance:

  • Migrated from poetry to uv for dependency management
  • Upgraded python from 3.8 to 3.13
  • Updated monai from 1.5.0 to 1.5.1
  • Updated flywheel-gear-toolkit to >=0.6,<0.7
  • Updated fw-utils from ^2.0.9 to >=5.1.0,<6
  • Migrated pyproject.toml to PEP 621 format
  • Updated Dockerfile to use multi-stage builds with python:3.13-slim base image
  • Added .dockerignore for faster Docker builds
  • Removed deprecated .env file in favor of .gitlab-ci.yml configuration
  • Updated .gitlab-ci.yml to use ci/gear.yml reference and configured medium runner with 3-hour timeout
  • Updated .pre-commit-config.yaml hooks to use sse reference
  • Added flywheel user (UID 31337) to Docker container
  • Configured pytest coverage threshold to 70%

Documentation:

  • Updated README with improved examples and removed deprecated AddChanneld usage
  • Updated example scripts to remove deprecated AddChanneld transform
  • Added new example int_norm_nifti.py for intensity normalization
  • Fixed markdown formatting issues in CONTRIBUTING.md

Breaking Changes:

  • User transform scripts using deprecated AddChanneld must be updated to EnsureChannelFirstd(keys, channel_dim="no_channel")

0.1.0 [2022-04-20]

Enhancements:

  • Added support for applying MONAI transforms to medical imaging files (NIfTI, DICOM, image formats)
  • Added user-defined transform script input for custom transformation pipelines
  • Added automatic validation of transform composition (must start with LoadImaged or LoadImage, end with SaveImaged or SaveImage)
  • Added support for both dictionary-based (LoadImaged) and non-dictionary (LoadImage) transform workflows

Documentation:

  • Added comprehensive README with usage examples and workflow diagram
  • Added CONTRIBUTING guide with dependency management, linting, testing, and release instructions
  • Added example transform scripts for common use cases (noise addition, cropping, training transforms)
  • Added release notes template

Maintenance:

  • Added CI/CD pipeline configuration with pre-commit hooks for code quality
  • Added test suite with coverage requirements (pytest, mypy, isort, pylint)
  • Added poetry for dependency management
  • Added Docker configuration based on python:3.8-buster
  • Added monai[nibabel,skimage,pillow,ignite,torchvision,itk,pandas, transformers,matplotlib,tifffile,imagecodecs] as core dependency
  • Added flywheel-gear-toolkit and fw-utils for Flywheel platform integration