Date(s) - 12/06/2012
PhD Oral Proposal
Over the past three decades Image Guidance (IG) has found ever-increasing applications in diagnostic and therapeutic procedures. While magnetic resonance imaging (MRI), positron emission tomography (PET), ultrasound, and x-rays all provide important tissue contrast, computed tomography with transmission x-rays provides the gold standard for high precision spatial targeting. Patient access during IG procedures often poses logistical problems. Access to two sides of the patient, one to facilitate entering and exiting x-ray beams, can often be problematic. When a procedure requires 3-dimensional anatomic information, as is necessary in operative guidance, the hundreds of x-ray views necessary for effective 3D reconstruction can often require 360-degrees of access.
Multiple alternative imaging techniques have been investigated to provide 3D anatomical data. One technique that shows increasing promise is Compton backscatter x-ray Imaging (CBI). This technique can potentially provide the spatial resolution of transmission imaging while only requiring access form one side of the patient. This imaging approach has the x-ray source and detector next to one another, on the same side of the patient, eliminating the need for capturing the exiting x-ray beam. Backscatter radiography has been used for industrial non-destructive testing for the last few decades. The current generation of CBI systems is optimized for common industrial use, such as air void detection in metal. These applications take advantage of large differences in atomic number and density between air and metal. In medical applications the difference between atomic number and density of two anatomical structures is not as great, and with medical imaging patient dose is a major concern.
To optimize CBI for medical imaging a better understanding of the underlying physics is necessary. Thus the first goal of this study is to develop a tool that will be used to characterize the two important system parameters: incident x-ray energy and detector geometry. Also, a mathematical model will be created based on radiation scatter and absorption principals to test the underlying radiation interaction principals. Because motion artifacts are a major concern when imaging animated biological subjects, it is essential to minimize the scan time. To this end, a simulated collimation system will be designed that is capable of obtaining data at multiple depths with a single view. Finally, a post-processing algorithm will be designed that enhances the raw output of the system. The signal will be enhanced by removing the noise contributed by scatter and absorption outside the region of interest but within the complex heterogeneous anatomical settings.