At the moment, the final csv file contains measurements (or rows) of all detected spots, and not only the ones that are kept after the nuclei-assignation and filtration, and these measurements have a nucleusNumber value = 0.
It would be nice to replace that with an empty value to avoid confusion where the user might think that the nuclei number starts with 0 and that these spots do belong to a nucleus.
In the meantime, excluding the rows that have a 0 value for nucleusNumber from the downstream analysis works well.