Automated Quantification of Multiple Cell Types in Fluorescently Labeled Whole Mouse Brain Sections Using QuPath.

Abstract

The quantification of labeled cells in tissue sections is crucial to the advancement of biological knowledge. Traditionally, this was a tedious process, requiring hours of careful manual counting in small portions of a larger tissue section. To overcome this, many automated methods for cell analysis have been developed. Recent advances in whole slide scanning technologies have provided the means to image cells in entire tissue sections. However, common automated analysis tools do not have the capacity to deal with the large image files produced. Herein, we present a protocol for the quantification of two fluorescently labeled cell populations, namely pericytes and microglia, in whole brain tissue sections. This protocol uses custom-made scripts within the open source software QuPath to provide a framework for the careful optimization and validation of automated cell detection parameters. Images obtained from a whole-slide scanner are first loaded into a QuPath project. Manual counts are performed on small sample regions to optimize cell detection parameters prior to automated quantification of cells across entire brain regions. Even though we have quantified pericytes and microglia, any fluorescently labeled cell with clear labeling in and around the nucleus can be analyzed using these methods. This protocol provides a user-friendly and cost-effective framework for the automated analysis of whole tissue sections.