Identifying primary objects not completely in full view

Hi all,

Is there a way to identify the nuceli of cells which aren’t completely in full view?
for example I have a couple of cells which are slightly off the edge of the image which I still want to include in the analysis

Hi @amandas,

Welcome to the forum.

There is a setting in IdentifyPrimaryObjects called “Discard objects touching the border of the image” that is likely set to Yes. If you switch this to No then these cells should be included (as long as what can be seen of them is within your diameter limits).

Just in case… if you are counting objects in a reduced number of fields to estimate the total population of a large sample, discarding all (or retaining all) border objects leads to biased results.
One traditional way to avoid this is to only count objects intersecting one (or both) of two boundaries (let’s say top and right) and ignore those in the left and bottom intersections (and do this for all the images/frames you acquired).

1 Like

I completely agree for, say, counting nuclei per field. However, would you agree that if you are counting sub-objects per nuclei (referencing their image), the best practice would be to remove all nuclei that are not fully in the field of view?

Yes, I would think so.