| The spatial arrangement of cells in the mammalian cortex directly relates to how the brain performs functionally. These cells include neurons, the fundamental building block of the neural network, and glia, which provide regulation, insulation, mechanical support, and nutrition for neurons and their processes. Currently, most analyses of pathological changes in spatial arrangement of neuron and glial cells rely on aberrations of cell placements large enough to be visible by the naked eye. We present a method that enables quantification of subtle spatial arrangement properties, or patterns, which exist right above biological noise and that may not be visually apparent due to the pattern's subtlety or lack of spatial cohesion. This method enables new comparisons between functional behavior of brain regions and quantitative measurements of detailed morphology within those regions. The method requires a three-fold effort of digitization, recognition, and analysis: we first develop the experimental platform needed to digitize tissue automatically at high resolution (1 microm per pixel) throughout whole tissue samples spanning the entire brain. We then develop the algorithms necessary to recognize cells within this digitized tissue, focusing on the challenge of delineating between neuron and glial cells in the same tissue sample. Finally, we show the analysis methods that are enabled by the new dataset. Among the methods are traditional measurements of total cell count and density, microcolumnar arrangement of neurons, and cross-correlation analysis between two cell populations such as neurons and glial cells. These measurements are either taken in known regions of the cortex or as "running windows" which show changes in spatial properties through the tissue irrespective of region delineations. The statistical robustness of the method is validated by comparing the results with models of multiple cell populations and changing spatial properties. We use the method to analyze cortical tissue samples from the rat and Rhesus monkey, and discuss current theories on mechanisms within the brain that may affect spatial arrangement of cells. Lastly, we describe future directions of study. |
