Mapping HPGe double-sided strip detector with x-rays for improved detector performance and position estimation

点击翻译按钮获取中文摘要

 DESCRIPTION (provided by applicant): Single Photon Emission Computed Tomography (SPECT) is a nuclear medicine imaging technique that allows for the mapping of the biological distribution of an injected radiotracer; depending on the type of radiotracer a range of topics may be studied, such as: mental and neurodegenerative disorders, various types of cancers, pulmonary embolisms, and heart disease. We are currently building a dual-headed SPECT system that uses double-sided strip, high-purity germanium (DSS HPGe) detectors. HPGe offers an energy resolution (FWHM < 1% at 140 keV) an order of magnitude better than conventional NaI gamma cameras, and its charge transport properties enable spatial resolutions similar to that of CdZnTe pixel detectors while using far fewer channels of readout electronics. The DSS configuration allows for sub-pixel positioning, however, multiple strip effects such as charge sharing, charge loss and Compton scattering within the detector can cause the events to be mis- positioned and/or uncounted. This decreases detection efficiency and overall sensitivity, possibly leading to artifacts in the reconstructed image. Alternative, advanced post-processing techniques have been shown to correct for these effects in limited cases (specific detector configurations and/or only small detector areas investigated). The goal of this proposed study is to investigate and implement the following post-processing techniques: waveform analysis, Maximum-Likelihood (ML) estimation, and charge loss correction across the whole detector. This will be done in three aims: Aim 1) Carefully measure the detector system response using the Advanced Photon Source (APS) at Argonne National Laboratory, Aim 2) Perform the position estimation techniques on the data obtained in aim 1. With these data sets (one for each technique) and knowledge of the beam position the three-dimensional position of each event may be determined using the ML estimation (note that one technique is to try the ML method without additional post-processing). Each technique will be evaluated by measuring: uniformity, energy and spatial resolution, and evaluation of a projected image (spatial resolution, signal- and contrast-to-noise ratio, modulation transfer function, etc), and Aim 3) Implement the new position estimation technique with the new SPECT system and evaluate the differences between the new reconstructed image compared to our current post-processing technique. We expect to obtain reconstructed SPECT images that have higher resolution, sensitivity, and contrast, as well as fewer artifacts, leading to better detectability.