/fuzzy-lasagna

Spatial mapping of human kidney by automated structural segmentation and registration across paired IHC, AF, and IMS datasets

Primary LanguageC++

fuzzy-lasagna

Spatial mapping of human kidney by automated structural segmentation and registration across paired IHC, AF, and IMS datasets

Download QuPath for analysis of whole-slide IHC images. You will also need the Bioformats Extension.
QuPath scripts are found in qupath_dev/:

  • .groovy scripts run as QuPath macros
    • Automate > Show Script Editor > File > Open...
    • open script, then Run
  • .js scripts are used in the ImageJ extension of QuPath
    • Extensions > ImageJ > Export region to ImageJ
    • in ImageJ: Plugins > Macros > Run

summary
Figure 1. Workflow and example output for automated MxIF segmentation using QuPath. A) Whole-slide tiling workflow as implemented in javascript macro. Each tile of image is processed for watershed maxima and intensity threshold in a single channel; resulting ROIs returned as QuPath annotations. B) Binary image mask outputs in three MxIF channels, highlighting unique macro structures in the tissue.

Python Implementation

pip install -r requirements.txt # make sure you have all necessary python packages