Automated Neurovascular Tracing and Analysis of the Knife-Edge Scanning Microscope Rat Nissl Data Set Using a Computing Cluster

Abstract

3D reconstruction of the neurovascular networks in the brain is a first step toward the analysis of their function. However, existing three dimensional imaging techniques have not been able to image tissues on a large scale at a high resolution in all three dimensions. For creating high-resolution neurovascular models, the Knife-Edge Scanning Microscope (KESM) at Texas A&M University has been developed and used to image whole rat brain vascular networks at submicrometer resolution.

In this thesis, I describe algorithms that are fully automatic and compatible with the large KESM rat Nissl data set. The method consists of image enhancement, binarization, 3D neurovascular networks tracing, and quantizing anatomical statistics. These methods are easily parallelizable and are compatible with high-throughput microscopy data. A computing cluster has been used to increase the throughput of the methods. Using the method developed, I analyzed a large volume of rat brain vasculature data. The results are expected to shed light on the structural organization of the vascular network that underlies the delivery of oxygen, nutrients, and signaling molecules throughout the brain.