A novel image segmentation method for the evaluation of inflammation-induced cortical and hippocampal white matter injury in neonatal mice

Abstract

The developing brain is very susceptible to environmental insults, and very immature infants often suffer from long-term neurological syndromes associated with white matter injuries such as periventricular leukomalacia. Infection and inflammation are important risk factors for neonatal brain white matter injuries, but the evaluation of white matter injury in animal models, especially the quantification of myelinated axons, has long been problematic due to the lack of ideal measurement methods. Here, we present an automated segmentation method, which we call MyelinQ, for the quantification of myelinated white matter in immunohistochemical DAB-stained sections of the neonatal mouse brain. Using MyelinQ, we show that a viral infection mimic agent, the Toll-like receptor 3 ligand Poly I:C, causes significant hypomyelination of white matter in the cortical and hippocampal fimbria regions, but not in the striatal caudoputamen region. We showed that MyelinQ can reliably produce results that are comparable to a method used in our previous publications. However, in comparison to the conventional method, MyelinQ has the advantages of being automated, objective and accurate. MyelinQ can analyze white matter in various specific brain regions and therefore provides a useful platform for the quantification of myelin and the evaluation of white matter injuries in animal models. MyelinQ and its code together with instructions for use can be found at: https://github.com/parham-ap/myelinq.