Skip to content

Add flexible label locations and ROI management#1

Open
TorecLuik wants to merge 2 commits intoNFDI4BIOIMAGE:mainfrom
Cellular-Imaging-Amsterdam-UMC:main
Open

Add flexible label locations and ROI management#1
TorecLuik wants to merge 2 commits intoNFDI4BIOIMAGE:mainfrom
Cellular-Imaging-Amsterdam-UMC:main

Conversation

@TorecLuik
Copy link

Add flexible label image locations and ROI management features 🤖✨

This PR enhances the Labels2Rois script to support more flexible workflows for comparing different segmentation methods on the same images.

Key Features Added: 🚀

  • Custom label suffixes: Configurable suffix parameter (default: -label) allows for varied naming conventions like _cellpose, _stardist, .0, etc.

  • Cross-dataset label search: New "Specific Dataset" mode enables storing label images in a separate dataset from target images via Label_Dataset_ID parameter

  • Named ROIs with method identification: ROIs are now named with suffix prefix (e.g., cellpose_1, stardist_2) making it easy to distinguish between different segmentation methods

  • Selective ROI clearing: Optional ROI management with filtering - clear all existing ROIs or only those matching specific text (e.g., remove only "cellpose" ROIs while preserving manual annotations)

Use Cases: 🔬

  • Compare CellPose vs StarDist segmentations on the same image
  • Iterative refinement workflows (auto + manual corrections)
  • Organized project structures with separate processing datasets
  • Method evaluation and benchmarking

Backward Compatibility: 🔄

All existing workflows continue to work unchanged. New features are opt-in via additional script parameters.

Files Changed: 📝

  • Labels2Rois.py: Core functionality enhancements
  • README.md: Documentation updates with examples and parameter reference

*🤖 Enhanced

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant