Robust Consensus Nuclear and Cell Segmentation

Melis Omer Irfan^1*Eduardo A Gonz Ález-Solares¹Tristan Whitmarsh¹Alireza Molaeinezhad¹Mohammad Al Sa'd¹Claire M Mulvey²Marta P Áez Ribes²Dario Bressan²Nicholas A Walton¹

• ¹University of Cambridge, Cambridge, United Kingdom
• ²Cancer Research UK Cambridge Institute, School of Clinical Medicine, University of Cambridge, Cambridge, England, United Kingdom

Cell segmentation is a crucial step in numerous biomedical imaging endeavors; so much so that the community is flooded with publicly available, state-of-the-art segmentation techniques ready for out of the box use. Assessing the virtues and limitations of each method on a tissue sample set and then selecting the optimum method for each research objective and input image is a time consuming and exacting task that often monopolizes the resources of biologists, biochemists, immunologists and pathologists; despite not being their project primary research goal. In this work, we present a segmentation software wrapper, coined CellSampler, which runs a selection of established segmentation methods and then combines their individual segmentation masks into a single optimized mask. This, so called 'uber mask', selects the best of the established masks across local neighborhoods within the image, where the neighborhood size and the statistical measure which determines the qualitative term 'best' are both chosen by the user.