Cell Well Segmentation v1.1

Cell Well Segmentation v1.1

First public release of Cell Well Segmentation, a desktop application for microscopy image cell segmentation, feature extraction, Manders colocalization analysis, QuPath-compatible GeoJSON export, and optional validation against ground-truth GeoJSON annotations.

This version provides a PyQt5 graphical interface designed to support batch processing of immunofluorescence microscopy images while preserving a reproducible output structure for downstream analysis.

Main features

• Graphical user interface for single-image and bulk-image processing.
• Support for TIFF, OME-TIFF and common raster image formats.
• Channel-based cell segmentation workflow for nuclei, red, green and cytoplasmic channels.
• ROI-based parameter exploration before running full-image processing.
• Instance mask export as TIFF.
• Cell-level feature extraction exported as CSV.
• Manders colocalization metrics at cell level.
• QuPath-compatible GeoJSON export for segmented cell regions.
• Fast bounding-box GeoJSON export mode for improved performance on large images.
• Preview figure generation for visual quality control.
• Existing-output handling with:
  • reprocess from zero,
  • skip completed outputs,
  • resume missing outputs.
• Optional validation using ground-truth GeoJSON annotations, including pixel-level DICE, IoU, precision and recall.

Main outputs

For each processed image, the software creates a dedicated output folder containing:

• instances.tif
• cell_features.csv
• manders_features.csv
• cell_features_with_manders.csv
• manders_summary.json
• qupath_final.geojson
• preview.png
• optional validation outputs under validation/

Notes

This is the first stable public release prepared for GitHub and Zenodo archival. The Zenodo DOI will be added after the release is archived.

Citation

If you use this software, please cite the GitHub/Zenodo release associated with this version.

DOI: To be added after Zenodo publication

Cell Well Segmentation v1.0

Cell Well Segmentation v1.0

First public release of Cell Well Segmentation, a desktop application for microscopy image cell segmentation, feature extraction, Manders colocalization analysis, QuPath-compatible GeoJSON export, and optional validation against ground-truth GeoJSON annotations.

This version provides a PyQt5 graphical interface designed to support batch processing of immunofluorescence microscopy images while preserving a reproducible output structure for downstream analysis.

Main features

• Graphical user interface for single-image and bulk-image processing.
• Support for TIFF, OME-TIFF and common raster image formats.
• Channel-based cell segmentation workflow for nuclei, red, green and cytoplasmic channels.
• ROI-based parameter exploration before running full-image processing.
• Instance mask export as TIFF.
• Cell-level feature extraction exported as CSV.
• Manders colocalization metrics at cell level.
• QuPath-compatible GeoJSON export for segmented cell regions.
• Fast bounding-box GeoJSON export mode for improved performance on large images.
• Preview figure generation for visual quality control.
• Existing-output handling with:
  • reprocess from zero,
  • skip completed outputs,
  • resume missing outputs.
• Optional validation using ground-truth GeoJSON annotations, including pixel-level DICE, IoU, precision and recall.

Main outputs

For each processed image, the software creates a dedicated output folder containing:

• instances.tif
• cell_features.csv
• manders_features.csv
• cell_features_with_manders.csv
• manders_summary.json
• qupath_final.geojson
• preview.png
• optional validation outputs under validation/

Notes

This is the first stable public release prepared for GitHub and Zenodo archival. The Zenodo DOI will be added after the release is archived.

Citation

If you use this software, please cite the GitHub/Zenodo release associated with this version.

DOI: To be added after Zenodo publication.