Publisher
The University of Arizona.Rights
Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author.Abstract
We have challenged traditional approaches to gamma-ray imaging and detection methods by exploring several novel ideas by both simulation and experiment. First, a gamma-ray simulation tool was developed to assist in the simulation of proposed ideas. The simulation tool can rapidly trace gamma rays with GPUs and also scintillation photons inside scintillation gamma-ray detector using multi-core CPUs. After that, a rotating-slit aperture design is re-explored with this simulation tool, based on this, a virtual pinhole synthesis (VPS) method is proposed, which can synthesize projection images similar to those of a physical pinhole, but with much higher sensitivity and resistance to noise. In addition, VPS only requires a 1-dimensional detector and enables adaptive imaging by making the slit adjustable. Next, the simulation tool was also used to explore a novel gamma-ray detector design, called the edge readout detector (ERD), which mounts light sensors on edge surfaces of a thin scintillator crystal to form a detector layer and stacks multiple detector layers to make a whole detector module. Both simulation and experiment results show that, by introducing optical barriers into it, the ERD can easily provide sub-millimeter resolution with direct DOI readout, good energy resolution and has the potential to reach < 500ps timing resolution. Next, a new detector calibration method called common data subset (CDS) was also proposed, which is capable of calibrating monolithic crystal detectors with fan beams. Complicated detector geometries can be calibrated eciently by CDS, including the edge-readout detector (ERD). Lastly, a new gamma-ray position estimation method called k-d tree search was explored. With the binary tree data structure and rules to prune branches while searching, the k-d tree search method can estimate interaction position rapidly, and 24 is also able to do position estimation accurately for detectors with complicated mean detector response functions (MDRFs).Type
textElectronic Dissertation
Degree Name
Ph.D.Degree Level
doctoralDegree Program
Graduate CollegeOptical Sciences