The ability to create large-scale high-resolution models of biological tissue provides an
excellent opportunity for expanding our understanding of tissue structure and function.
This is particularly important for brain tissue, where the majority of function occurs at the
cellular and sub-cellular level. However, reconstructing tissue at sub-cellular resolution is
a complex problem that requires new methods for imaging and data analysis.
In this dissertation, I describe a prototype microscopy technique that can image large
volumes of tissue at sub-cellular resolution. This method, known as Knife-Edge Scanning
Microscopy (KESM), has an extremely high data rate and can capture large tissue samples
in a reasonable time frame. We can therefore image complete systems of cells, such as
whole small animal organs, in a matter of days.
I then describe algorithms that I have developed to cope with large and complex data
sets. These include methods for improving image quality, tracing filament networks, and
constructing high-resolution anatomical models. These methods are highly parallel and designed
to allow users to segment and visualize structures that are unique to high-throughput
microscopy data. The resulting models of large-scale tissue structure provide much more
detail than those created using standard imaging and segmentation techniques.
Identifer | oai:union.ndltd.org:tamu.edu/oai:repository.tamu.edu:1969.1/ETD-TAMU-2009-05-745 |
Date | 16 January 2010 |
Creators | Mayerich, David |
Contributors | Keyser, John |
Source Sets | Texas A and M University |
Language | en_US |
Detected Language | English |
Type | Book, Thesis, Electronic Dissertation |
Format | application/pdf |
Page generated in 0.0103 seconds