An Automated Analysis Pipeline for Microglia Morphology in Nikon NIS-Elements

The morphology of a microglial cell determines its activation state and is indicative of various pathologies. Mouse brains are composed of 300,000-500,000 microglia, which makes the application of automated measurement an attractive method for activation status quantification. Additionally, automation can allow for the maximal amount of data per animal versus manual methods. Here we describe an annotated analysis pipeline that works with the commonly used analysis software package Nikon NIS-Elements to automate quantification of microglia cells in mounted and stained brain sections given user-specified parameters. The pipeline can be modified for use with other types of ramified cells, such as neurons. Our analysis pipeline includes pre-processing, thresholding, skeletonization, puncta detection, Sholl analysis, branch classification, and measurement. Once detection thresholds for the desired cell population are validated on representative fields of view, batch analysis can be run for as many image files as desired to generate tables of measurements for regions as large as entire microscope slides when run on a dedicated image processing workstation. The pipeline can also be customized for other purposes, unlike some other automated analysis tools. To our knowledge, a publicly available pipeline template does not exist that can process this amount of microglial morphology and yet is modular, utilizing the customizable NIS-Elements General Analysis 3. With our pipeline, analysis is fast and hands-off once initial thresholds are set for segmentation, introducing less human bias compared with manual measurements, generating more useful data per animal, and allowing less time to be spent on analysis and more time on experiments.