Doctor of Philosophy (PhD)
Physics and Astronomy
Several areas of prostate cancer (PCa) management, such as imaging permanent brachytherapy implants or small, aggressive lesions, benefit from high image resolution. Current PCa imaging methods can have inadequate resolution for imaging these areas. Endorectal digital prostate tomosynthesis (endoDPT), an imaging method that combines an external x-ray source and an endorectal x-ray sensor, can produce three-dimensional images of the prostate region that have high image resolution compared to typical methods. This high resolution may improve PCa management and increase positive outcomes in affected men.
This dissertation presents the initial development of endoDPT, including system design, image quality assessment, and examples of possible applications to prostate imaging. Experiments using computational phantoms, physical phantoms, and canine prostate specimens were conducted.
Initial system design was performed computationally and three methods of endoDPT image reconstruction were developed: shift and add (SAA), backprojection (BP), and filtered BP (FBP). A physical system was developed using an XDR intraoral x-ray sensor and a GE radiography unit. The resolution and radiation dose of endoDPT were measured and compared to a GE CT scanner. Canine prostate specimens that approximated clinical cases of PCa management were imaged and compared using endoDPT, the above CT scanner, and a GE MRI scanner.
This study found that the resolution of endoDPT was significantly higher than CT. The radiation dose of endoDPT was significantly lower than CT in the regions of the phantom that were not in the endoDPT field of view (FoV). Inside the endoDPT FoV, the radiation dose ranged from significantly less than to significantly greater than CT. The endoDPT images of the canine prostate specimens demonstrated qualitative improvements in resolution compared to CT and MRI, but endoDPT had difficulty in visualizing larger structures, such as the prostate border.
Overall, this study has demonstrated endoDPT has high image resolution compared to typical methods of PCa imaging. Future work will be focused on development of a prototype system that improves scanning efficiency that can be used to optimize endoDPT and perform pre-clinical studies.
Steiner, Joseph Robert, "Endorectal Digital Prostate Tomosynthesis" (2018). LSU Doctoral Dissertations. 4582.