Charting out the complete brain microstructure of a mammalian species is a grand challenge. Recent advances in serial sectioning microscopy such as the Knife- Edge Scanning Microscopy (KESM), a high-throughput and high-resolution physical sectioning technique, have the potential to finally address this challenge. Nevertheless, there still are several obstacles remaining to be overcome. First, many of these serial sectioning microscopy methods are still experimental and are not fully automated. Second, even when the full raw data have been obtained, morphological reconstruction, visualization/editing, statistics gathering, connectivity inference, and network analysis remain tough problems due to the unprecedented amounts of data. I designed a general data acquisition and analysis framework to overcome these challenges with a focus on data from the C57BL/6 mouse brain. Since there has been no such complete microstructure data from any mammalian species, the sheer amount of data can overwhelm researchers. To address the problems, I constructed a general software framework for automated data acquisition and computational analysis of the KESM data, and conducted two scientific case studies to discuss how the mouse brain microstructure from the KESM can be utilized. I expect the data, tools, and studies resulting from this dissertation research to greatly contribute to computational neuroanatomy and computational neuroscience.

The Knife-Edge Scanning Microscope (KESM) enables imaging of an entire mouse brain at sub-micrometer resolution. By using the data sets from the KESM, we can trace the neuronal and vascular structures of the whole mouse brain. I investigated effective methods for automatic seedpoint selection on 3D data sets from the KESM. Furthermore, based on the detected seedpoints, I counted the total number of somata and traced the neuronal structures in the KESM data sets. In the first step, the acquired images from KESM were preprocessed as follows: inverting, noise filtering and contrast enhancement, merging, and stacking to create 3D volumes. Second, I used a morphological object detection algorithm to select seedpoints in the complex neuronal structures. Third, I used an interactive 3D seedpoint validation and a multi-scale approach to identify incorrectly detected somata due to the dense overlapping structures. Fourth, I counted the number of somata to investigate regional differences and morphological features of the mouse brain. Finally, I traced the neuronal structures using a local maximum intensity projection method that employs moving windows. The contributions of this work include reducing time required for setting seedpoints, decreasing the number of falsely detected somata, and improving 3D neuronal reconstruction and analysis performance.