Optimizing the system parameters of BS-SPECT for small animals imaging

碩士 === 國立清華大學 === 核子工程與科學研究所 === 101 === Single photon emission computed tomography (SPECT) employing pinhole collimators are capable of generating images with sub-millimeter spatial resolution. However, the pinhole collimator is expensive, heavy and difficult to fabricate. In our previous study, be...

Full description

Bibliographic Details
Main Author: 沈承緯
Other Authors: 莊克士
Format: Others
Language:en_US
Published: 2013
Online Access:http://ndltd.ncl.edu.tw/handle/41302247332971434337
Description
Summary:碩士 === 國立清華大學 === 核子工程與科學研究所 === 101 === Single photon emission computed tomography (SPECT) employing pinhole collimators are capable of generating images with sub-millimeter spatial resolution. However, the pinhole collimator is expensive, heavy and difficult to fabricate. In our previous study, beam stopper (BS) device was employed to replace the pinhole collimator for high resolution SPECT imaging. In BS-SPECT system, dual scans with and without the BS are conducted for all directions, and the difference between these two sinograms yields the pinhole equivalent projections. For optimization, a novel ray-tracing model was proposed to evaluate the system performance of BS-SPECT system in this study. By omitting the repeated random sampling, ray-tracing simulations are more efficient than Monte Carol simulations. This model assumes that the BS-SPECT system consists of a flat detector combined with a BS. By calculating the point spread functions (PSF's) of various system geometries and BS designs, the total sensitivities and resolutions can be derived. The results show that (1) circle BS made of gold has the optimal system performance; (2) high sensitivity-to-FWHM ratio can be achieved by using large BS; (3) the BS should be placed as close as possible to the object for optimal sensitivity and resolution; (4) the detector can be placed away from the object to improve the resolution without sensitivity losses; (5) the highest SRN is derived by using 1:1 scan time ratio.