README.md

metadata

title: ZO-1 Network Analysis (RIS + TiJOR)
emoji: 🔬
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: 5.44.1
app_file: app.py
python_version: 3.11
tags:
  - gradio
  - bioimage
  - segmentation
  - cellpose
  - computer-vision

ZO-1 Network Analysis Tool

Gradio app for AI-powered segmentation and quantification of ZO-1 tight junction networks using:

• RIS (Radial Integrity Score): concentric circle crossings per pixel length
• TiJOR (Rectangular method): rectangle perimeter crossings per pixel length

How to use

1. Upload a ZO-1 image (TIFF, PNG, JPG). 16-bit TIFFs are supported.
2. In Segmentation:
   • Adjust Cell Diameter Estimate (px) if results are off
   • Scale Factor 1.0 = full size (lower if slow)
   • AI Contour Validation can improve contours but is slower
3. In Analysis:
   • Pick RIS (circles) or TiJOR (rectangles)
   • Tune parameters (radii/rectangles, min separation)
   • Optionally show contours/geometry/cross-sections
4. In Export:
   • Download CSV or Text report; files are named <image>_RIS.csv/.txt or <image>_TiJOR.csv/.txt

Features

• Cellpose-based segmentation (GPU or CPU)
• Robust TIFF handling (8/16-bit)
• RIS and TiJOR analyses with visual overlays
• Exports with image-based filenames

Parameters

• Segmentation: Cell Diameter (px), Scale Factor, AI Validation
• RIS: κ (packing factor), min/max radius (%), number of circles, min separation (px)
• TiJOR: initial/max rectangle size (%), steps, min cross-section distance (px)

Notes

• Large images can be slow; reduce Scale Factor if needed
• GPU is recommended for faster segmentation

Acknowledgements

• Built with Gradio and Cellpose