I don’t know for the recent update, but at least until now following guidelines have proven useful for me:
Analysis with CP will be easier if you have fluorescence-based images (images that have a bright foreground and dark background), generally speaking.
While it is possible to quantify a wide range of features related to size, shape, orientation, or distance between them, your greatest problem will be detecting the nuclei (and cells) from the phase contrast image alone.
You could try identifying the nuclei only (e.g. DAPI stain) and measure the distance between them, which may resemble the increase in size/length (as they appear to remain confluent). This would be considerably simpler than the following:
Alternatively, you could try the bundled version if ilastik (a machine-learning tool) that comes with CP. It will take some time to understand the features, but you could use this software to detect the cells (and maybe nuclei), and export a file that lets CP detect them too. Check out the help for the module “ClassifyPixels”.
p.s. Your cells look endothelial If they are, CD31 staining will wonderfully highlight the intercellular junctions, although the localisation may change upon TNF-stimulation. Should still work nicely though.