Counting cells manually in 3D with ImageJ
Counting cells or nuclei automatically in 3D remains elusive because in so many biological systems they are too close together or have complex morphologies so segmenting robustly does not work.
Therefore, we count manually. We do not have tools to reach into a spatial representation of our data. Therefore, we do counting in 2D.
It helps to have tools to compress the 3Dness and to mark features.
This is one such tool.
As you move through Z axis, you click on features (or us the F5 function key) to mark and count them. These counts are overlaid on the image. The markings can be turned on and off.
Also, Z series can be compressed in Z. For instance, a Z stack with 60 images could be compressed by a factor of 4 to 15 slices which is more magnageable to traverse and count without having featues directly above/below each other combine.
These macros are one way of doing this. They could be bundled with other macros you like to use. They may be modified if you find them helpful but have a variation of workflow or metrics you prefer.
A very brief video tutorial is also included.
Manually counting does have a benefit: it forces us to look at the data and see features and relationships automated counting misses.
If you are maually counting 2D images, the cell couunter built in to Fiji is probably better than this tool. Also, we should consider revisiting strategies for segmentaing automatically.
Other favorite macros here.
last updated 2024-02-03