Font Size: a A A

Image Reconstruction Optimization and Quantification for Image-Guided Luminescent Tomography

Posted on:2015-01-12Degree:Ph.DType:Dissertation
University:Dartmouth CollegeCandidate:Holt, Robert WilliamFull Text:PDF
GTID:1478390017496483Subject:Physics
Abstract/Summary:
Molecular imaging of cancer features is a critical part of advancing better tools for oncology management and drug discovery, yet it is still evolving as a useful tool. Diffuse Fluorescence Tomography (FT) is one approach to molecular imaging, used for excitation and detection of emitted light signals, which uses spatial reconstruction of the location of the fluorescence distribution to map a selected molecular species within biological tissue. These maps can provide a powerful in vivo tool to monitor molecular processes of health and disease or predict efficacy of targeted cancer therapies. The injected fluorescence tracers used in FT emit light at near-infrared wavelengths where absorption by tissue is low and so imaging through many centimeters of tissue is possible. Applications involve use of targeting biological molecules, such as cancer receptors overexpressed on cell surfaces. Specific targeted ligands injected as intravenous tracers then ideally show where the corresponding receptor molecules are present through in vivo binding and localization.;The main goal of this project was to move toward a clinically relevant, automated, quantitative form of FT imaging, with multiple independent advances to the technique. First, a method of geometric system optimization is proposed and confirmed based on establishing a uniform detection sensitivity map. Second, analysis of data processing techniques and time domain methods show that time domain imaging techniques can improve imaging accuracy while lowering imaging time, but is largely not worth the additional cost and does not provide as much improvement as structural hybrid imaging. A review of previously proposed anatomical prior information was performed, followed by a proposed technique which automates the inclusion of anatomical prior data by the construction of a regularization functional based on an anatomical image. In an application example, multispectral techniques are used to resolve more than one tracer simultaneously, which allows quantification of a tracer targeted to the epidermal growth factor receptor on cancer cells. Finally, in an application of these tomographic methods in radiation oncology, a method is demonstrated to measure oxygenation of human brain tumors during external beam radiation therapy using an oxygen sensitive phosphorescent agent with FT recovery of the signal.
Keywords/Search Tags:Imaging, Cancer
Related items