(part 1) Step 0: auto ROIs generation

For this step, make sure you installed multi-template matching plugin (see the link, it also requires IJ-OpenCV update site) for Fiji.

1. Open your multi-channel z-stack in Fiji and duplicate only pan-ExM channel (make a new window).

2. Select a z-slice where the majority of basal bodies cross-sections look like a donut.
   For example, this is a good slice:
   
   and this is a bad slice selection:
   

3. Write down or memorize the slice number. Duplicate (create new image) with only that slice (called single_slice.tif here).

4. On the slice, choose a most representative image of basal body donut, select it with rectangular ROI and extract it to the new image (duplicate). Let's rename this image to single_donut.tif for clarity.
   

5. Run Plugins -> Multi-Template-Matching -> Template Matching Image command. In the new dialog, specify:

   • Template field should be single_donut.tif (or whatever you called it);
   • Image field should be single_slice.tif (or whatever you called it);
   • Matching method selection is not so important, we want approximate detection;
   • Number of objects expected should be approximately equal to the number of basal bodies. You can iterate this process and try different values;
   • Add ROI detected to ROI manager field should be checked.

6. In the end it should produce a set of rectangular ROIs detections added to the ROI manager.

7. At this point you can add/delete/edit ROIs of detected basal bodies.
8. Run s0_rename_ROIs_set_position.ijm macro (in the part1_cells_processing folder). It will ask you to provide the slice number (from the step 2-3).
9. Now all ROIs' names in the ROI manager should be in the number format and positioned at the proper slice. Select all ROIs in ROI manager and save them on disk as zip file. We are going to use them in the next step.

Next is (part 1) Step 1: detecting basal bodies (BB) in 3D