ImageJ1 macros to run Cellpose within Fiji (using the BIOP-EPFL plugin) and to match label values in independent labelmaps of nuclei and cell bodies.
NOTE: These macros will not work with the latest version of the BIOP Cellpose Wrapper! In the comings weeks we will (hopefully find the time to) update this repository.
ImageJ1 macro to run Cellpose on an image or active (rectangular) selection and display the segmentation as colored overlay. Labels are also added to the ROI Manager.
Input: 2D image or 2D timelapse.
Output: (timelapse) labelmap, ROIs
Requirements:
- A working Cellpose Python environment.
- Fiji Update Site: PTBIOP
- Follow the Enable conda command outside conda prompt part of the instructions on https://github.com/BIOP/ijl-utilities-wrappers
Note that the hard work has been done by others, namely the Cellpose authors and the BIOP at EPFL.
Please cite the Cellpose article if you use this macro in a publication.
N.B. In this macro the 'cyto2' model currently does not work on 2-channel images (cells and nuclei). This should be fixed very soon.
Macro to match the label values of a nuclei labelmap with a cell body labelmap. Cells can have multiple nuclei. Nuclei without a cell body are removed.
Input: Two 2D labelmaps.
Output:
- labelmap with nuclei labels reassigned to match the cells
- labelmap of the cytoplasm (cell body minus nuclei)
- labelmap of cells without an assigned nucleus (empty cells)
- labelmap of cells without an assigned nucleus minus any other nuclei that overlap, but are assigned to another cell
- (Not yet available) labelmap of nuclei without a cell body
Requirements:
- Fiji Update sites: CLIJ and CLIJ2
Please cite the CLIJ article if you use this macro in a publication.
Input label images: nuclei and cells (label values do not match - colors are different):
Output label images: reassigned nuclei, cytoplasm, empty cells with nuclei overlap removed:
