| Abstract: | The vascular architecture of the brain is very complex and reconstruction of this architecture can aid in understanding the functions of vessels in different regions of the brain. Advances in microscopy have enabled high-throughput imaging of massive volumes of biological microstructure at a very high resolution. The Knife Edge Scanning Microscope (KESM), developed by the Brain Network Laboratory at Texas A & M University, is one such instrument that enables imaging of whole small animal brains at sub-micrometer resolution. The KESM has been successfully used to acquire vasculature dataset from a mouse brain stained by India ink. However, manual tracing and reconstruction of vessels is not feasible due to the huge volume of the dataset. Therefore, developing efficient and robust automatic tracing methods is essential for analysis of the network. This thesis presents an efficient skeletonization based tracing algorithm to reconstruct the vascular structure in the KESM India ink data set. The skeleton is generated from the original volume by repetitively removing voxels from the object's boundary such that the connectivity and topology in the original volume is preserved. The skeleton is then dilated to reconstruct the volume. The accuracy of this method is determined by comparing the reconstructed volume with the original volume. This method is expected to trace the entire vascular network of the brain quickly and with high accuracy without human assistance. The electronic version of this dissertation is accessible from http://hdl.handle.net/1969.1/155652 |
|---|