Separating clumping cells



I´m working with a cell type that forms cellular clusters in the cultures (representative image attached). So far I have tried to create a pipeline that would separate the nuclei (and debris attached to the clusters), in order to count e.g. MAP-2/b-tubulin positive cells in the in the cultures. I have not been very successful, but is it actually possible to declump single cells by image analysis if the growth pattern is as challenging as this? Any good tips for this?

Thank you for any help!