How about confocal z-sections?



Hi all,

Currently my image sets are confocal collapsed stacks.
Can CP detect objects from individual z-sections, and determine an exact number of objects within the z space?
This may be difficult, because I don’t want CP to count an object twice, if the object is within 2 sections.

So in the end, CP will output a total number of objects per cell…

Can this be done?

Thanks heaps



Hi Tim,

As of this point, CellProfiler doesn’t handle Z-stacks directly. But this functionality might be approximated using TrackObjects. Provided (a) your objects are reasonably well-defined from Z-slice to Z-slice, and (b) you feed in the images in Z-order, I think TrackObjects will serve to identify objects that are continuous in Z-space. However, TrackObjects is in a very beta form in the latest compiled release (although we’ve done a lot of development on it since then).

If that doesn’t work, and your objects are not stacked on top of one another, you might use CorrectIllumination_Calculate without smooth to create a averaged projection of the Z-series, which can then be segmented for your object count.

Alternately, this projection can be used as a binary mask, and then re-loaded with LoadSingleImage into a second pipeline to mask each Z-slice in turn, Relate the objects found in the slice as children to the “parent” objects in the mask, and perhaps distinguish the 3-D objects by which 2-D parent object in the mask they match up with.

There are probably additional ways to approach this issue, but it’s a start!