Feature-Based Analysis of Microvasculature in High Resolution Microscopy Images of Mice Brains



Journal Title

Journal ISSN

Volume Title



Quantitative analysis of three dimensional image datasets in microscopy has the potential to assist histopathology researchers in understanding anatomy, in identifying the cause of diseases or disorder, in the development of new methods in medicine. This project presents steps for tracing, quantification, and analysis of mouse brain microvasculature in datasets obtained from a high throughput microscopy technique called Knife-Edge Scanning Microscopy (KESM). Identification and tracing of microvessels is achieved by segmenting the microvasculature from the surrounding tissue into binary images, obtaining the microvessel centerlines by thinning the segmented microvasculature using binary three dimensional structuring elements, then determining the unit width voxel curve skeleton from the result. This method overcomes computational limitations by forming a computational structure that is easily parallelized on disjoint blocks of data. The unit width skeleton is used to form the nodes of a graphs from which statistics, such as segment length, curvature, and orientation of segments, average branches in a network are computed.



Skeleton, Tracing, Brain, Microvessels