Skip to content

Release Notes

0.1.2_4.0.6 [2025-12-16]

Maintenance:

  • Updated urllib3 to >=2.6.0 to address security vulnerabilities
  • Updated brotli to >=1.2.0 to address security vulnerabilities

0.1.1_4.0.6 [2025-10-28]

Enhancements:

  • Added Cellpose gear for cell and nuclei segmentation in microscopy images
  • Implemented support for multiple Cellpose models (cyto, nuclei, cyto2, transformer, omni)
  • Added 3D segmentation support for volumetric images with configurable z-axis and channel axis
  • Implemented GPU acceleration support with automatic CPU fallback
  • Added cell statistics output with area and centroid coordinates in CSV format
  • Created visualization outputs with segmentation masks and overlays
  • Pre-downloaded Cellpose-SAM transformer model weights from Hugging Face during build

Maintenance:

  • Migrated to uv for dependency management
  • Updated to python 3.12 runtime
  • Added cellpose 4.0.6 with pytorch 2.6.0 and CUDA 12.6 support
  • Added flywheel-gear-toolkit >=0.6,<0.7 for Flywheel integration
  • Added numpy >=1.22.0,<2.0, scipy >=1.9.0,<2.0, and pandas <2.0
  • Added scikit-image >=0.19.0,<0.22, opencv-python-headless >=4.7.0,<5.0, and pillow 11.3.0
  • Added matplotlib >=3.5.0,<4.0 for visualization
  • Added huggingface-hub 0.35.3 for model weight management
  • Configured GitLab CI with 6-hour timeout and GPU support
  • Set up pre-commit hooks with ruff, hadolint, jsonlint, yamllint, and pytest

Documentation:

  • Added comprehensive README with usage instructions, workflow diagrams, and FAQ
  • Added CONTRIBUTING guide with development setup and testing instructions
  • Included citation for Cellpose paper and BSD-3-Clause license