Quantitative Investigation Of Brain-wide3D Detailed Vascular Structure In The Mouse Brain Based On Cytoarchitecture

Our brain is the most important and complex organ. However, we still understand little about the inner detailed structure and working mechanism. The brain consists of highly organized and interconnected neurons, glia cells, and blood vessels. It is of great importance to investigate the detailed architecture for the understanding of working mechanism of brain function, the signal interpretation of functional brain imaging, the understanding of some brain diseases and development. However, mainly limited by the development of labelling, imaging and image processing techniques, the vascular-cellular configurations analysis in the mammalian (e.g., mouse) brains were confined in the cerebral cortex, leaving the brain-wide detailed vascular-cellular configurations largely uncharted. Here, this thesis constructs a systematic quantitative analysis method of detailed vascular-cellular configurations in the brain-wide scale, and applies the method to characterize the vascular-cellular configuration of normal adult mouse brains in the level of cortical columns, brain regions, and the whole brain.The original whole mouse brain datasets were obtained using micro-optical section-ing tomography. The blood vessels and cells in the mouse brains were labelled using Nissl staining and indian-ink perfusion. For the vascular vectorization, three key technical prob-lems were tackled:①Utilizing iterative tracing strategy to reduce the sensitivity of noise and uneven background intensity, which is a major problem of traditional skeleton method.②Dropping the hypothesis of shape and size for accurate tracing of both big vessels and capillaries, which is a big challenge of model fitting methods.③A building-block design for rapid and robust tracing. During tracing, the modules were automatically selected ac-cording to the estimated diameter. The big vessels that have high signal-to-noise ratio were traced using rapid module, and the microvessels that have low signal-to-noise ratio were traced using robust module. The cells were vectorized using FARSIGHT software. Based on the vectorization results, a quantitative analysis software was designed to compute four parameters:vascular length density, vascular fractional volume, cell number density, and the distances from cells to the nearest blood vessel. To overcome the speed problem of whole brain analysis, a segment-based data structure was designed for fast indexing in a sliding window.Based on the established techniques, the detailed vascular-cellular configuration were investigated in the level of cortical columns, brain regions, and the whole mouse brain. In the level of cortical columns, the specificity and similarity of vascular-cellular distribution were characterized in the barrel cortex. In the layer Ⅳ, the penetrating vessels and microvessels were found to have a columnar distribution, while the branches of penetrating vessels have inter-columnar connection and selective connection of neighbouring columns. The direct3D reconstruction and quantitative analysis results are different with a traditional model and the recent results of a research group, suggesting a much more complex vascular distribution among cortical columns. In the level of the brain-wide brain regions, the cells and blood vessels in six representative brain regions (prefrontal cortex, primary motor cortex, barrel cortex, primary visual cortex, striatum, and amygdala) were vectorized and analysed. The analyses of laminar distribution show that the variation in vascular density is significantly higher in the middle layers than in the upper and lower layers. The results of parameter value distribution show that the distances from cells to the nearest blood vessel is stable, suggest-ing a steady energy supply. The results of correlation analysis among the parameters show that the distance from cells to the nearest blood vessel is linearly correlated with vascular length density, which may benefit the data interpretation of vascular imaging. In the level of the whole mouse brain, the blood vessels were vectorized for the first time, and a vascular length density map were built, which may have potential application for the study of brain metabolism and functional brain imaging.In conclusion, this thesis established a quantitative analysis method for the3D brain-wide and detailed vascular-cellular configuration, and applied in several normal adult mice. This integrated method provides a systematic overview of the detailed3D vascular and cel-lular configuration, which opens up an avenue to explore the enigma of the vascular and cellular configurations in normal, developing and diseased mammalian brains.