Comparisons of quantitative approaches for assessing microglial morphology reveal inconsistencies, ecological fallacy, and a need for standardization

Microglial morphology is used to measure neuroinflammation and pathology. For reliable inference, it is critical that microglial morphology is accurately quantified and that results can be easily interpreted and compared across studies and laboratories. The process through which microglial morphology is quantified is a key methodological choice and little is known about how this choice may bias conclusions. We applied five of the most commonly used ImageJ-based methods for quantifying the microglial morphological response to a stimulus to identical photomicrographs and individual microglial cells isolated from these photomicrographs, which allowed for direct comparisons of results generated using these approaches. We found a lack of comparability across methods that analyzed full photomicrographs, with significant discrepancies in results among the five methods. Quantitative methods to analyze microglial morphology should be selected based on several criteria, and combinations of these methods may give the most biologically accurate representation of microglial morphology.