Maximum Likelihood Solution of Arrival Time for the Event Pulse Modeled by Filtered Inhomogeneous Poisson Process in PET Detectors

Zhenzhou Deng,Yuqing Liu,Zhentong Wu,Qingguo Xie
2015-01-01
Journal of Nuclear Medicine
Abstract:1842 Objectives Timing has become an absorbing technique for PET detectors since time-of-flight (TOF) PET has shown its potential to significantly improve the image quality. If a 500 picoseconds (ps) full-width at half maximum (FWHM) coincidence timing resolution (CTR) can be achieved in a PET system, many benefits can be realized for whole-body FDG imaging[1-4], including a) lessening of axial blurring; b) better random event rejection; c) simultaneously acquiring emission and transmission data; d) reduction of the required dose and scanning time; e) improvement in image spatial resolution; f) recovering low input contrast; g) overcoming inconsistent or missing data. In this study, we develop a time pickoff method to obtain a better time resolution. Methods Signals in PET detectors or event pulses are always generated by a two-step transformation: A. Scintillators convert the Gamma photons of 511 keV into soft visible and ultraviolet light fluxes. B. Photomultipliers convert the light fluxes into electrical pulses. Considering that the emitted scintillation photons follow Inhomogeneous Poisson statistics, and effect of photoelectric transformation is approximately equivalent to a convolution, we modeled an event pulse as Filtered Inhomogeneous Poisson (FIP) Process. Event time is picked up when a time point satisfies Maximum Likelihood (ML) criterion. Results Timing spectrum produced by leading edge discriminator (LED), constant fraction discriminator (CFD) and energy normalization-based leading edge discriminator (EN-LED) and ML are shown in the figure. The CTR are 187.35 ps, 190.96 ps, 165.16 ps and 151.81 ps, respectively. Conclusions In this work, we have proposed a PET event timing pickoff method based on FIP model. The ML solution of arrival time aims at maximizing the likelihood of the time difference to be estimated with the given samples. Experimental results showed the ML method has better performance to retrieve timing information. Research Support The authors thank Mr. Ming Niu and Mr. Jun Zhu for assisting the experiment setup, the PET center of Union Hospital for providing the radioactive sources, Ms. Jia Hao for providing the LaBr3:Ce crystals.
What problem does this paper attempt to address?