Detector Concepts for Time-of-Flight PET

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): PET imaging performance fundamentally relies on the performance of the detector. In recent years detectors with very good timing performance have been incorporated into time-of-flight (TOF) PET scanners by each of the major manufacturers (Philips, Siemens, GE) who promote TOF PET/CT instruments as their top-of-the-line products. These instruments achieve 500-600 ps TOF resolution, and achieve very good performance by building upon many advances in instrumentation from the last decades, including high-resolution detector encoding schemes, stable electronics, powerful computers, accurate data correction and iterative image reconstruction algorithms (with modeling) for 3D acquisition, and integration of CT for attenuation correction and anatomic registration. It is widely held that TOF has made an impact on image quality and that it provides a clinical benefit for whole-body imaging, in particular oncology studies where lesion quantification and detection are critical. It is understood that improved TOF resolution should lead to further improvements, although the exact relationship between timing resolution and metrics to characterize clinical tasks is complicated. Nevertheless, we believe that improved timing resolution will translate to improved PET imaging performance, and is likely to provide further benefit for both clinical and research investigations. In this research proposal we seek to develop and evaluate a detector design with superior timing resolution compared to designs in practice today. We do recognize that improved detector timing resolution must be weighed against other factors of detector performance, in particular spatial resolution and sensitivity. In fact, early TOF PET scanners in the 1980's were eventually phased out since they could not compete against the better spatial resolution and higher sensitivity of conventional PET scanners at the time. Therefore, we will consider the impact of improved timing resolution together with other detector performance metrics on the overall performance of the system. We will also develop a design for TOF that also includes the ability to discriminate depth-of-interaction (DOI). This design may trade-off ultimate timing performance, but will improve system spatial resolution by reducing parallax error, and allow for the design of whole-body systems with smaller ring geometry. The specific aims include 1) develop detector concepts to improve timing resolution with lanthanum bromide scintillators and silicon photo-multipliers, 2) explore ways to measure depth-of-interaction (DOI) while also preserving timing performance, 3) develop proto-type detector arrays to demonstrate improved timing resolution in a practical configuration with spatial resolution and sensitivity appropriate for PET applications, and 4) illustrate the impact of improved detectors using computer models and clinical imaging metrics, such as lesion contrast and detectability. At this project's conclusion we will have developed and characterized a new detector with improved performance for TOF PET that combines fast scintillators with solid-state photosensors.