Towards polarization-enhanced PET: Study of random background in polarization-correlated Compton events

TL;DR

This paper investigates polarization-enhanced PET by leveraging polarization entanglement between annihilation photons and detecting their Compton-scattering-induced polarization correlations. It implements two GAGG:Ce detector modules read out by SiPMs with a TOFPET2 system to measure coincidence events from a Ga-68 source, extracting the polarimetric modulation factor μ from acceptance-corrected azimuthal differences via a Pryce-Ward-like relation. The key finding is that polarization-correlated (Compton) events exhibit higher signal-to-random-background (SBR) ratios than conventional single-pixel events, with SBR improving as μ increases, suggesting μ as an auxiliary estimator of random background. These results support the potential of polarization-sensitive PET to enhance background suppression in clinical imaging and may influence polarization-based approaches to Positronium imaging in PET contexts.

Abstract

Positron Emission Tomography (PET) is a medical imaging modality that utilizes positron-emitting isotopes, such as Ga-68 and F-18, for many diagnostic purposes. The positron annihilates with an electron from the surrounding area, creating two photons of 511 keV energy and opposite momenta, entangled in their orthogonal polarizations. When each photon undergoes a Compton scattering process, the difference of their azimuthal scattering angles reflects the initial orthogonality of their polarizations, peaking at $\pm$90$^{\circ}$. This type of correlation is not yet utilized in conventional PET scanners, but could potentially offer an energy-independent method for background reduction. Measurements of these kinds of correlations can be achieved using Compton polarimeters, built from a single layer of segmented scintillating crystals such as Gadolinium Aluminium Gallium Garnet doped with Cerium (GAGG:Ce), read out by silicon photomultipliers (SiPMs). In this paper, we study the signal-to-random background ratios in measurements of these correlated annihilation photons from coincidence time spectra across clinically relevant source activities, from $\sim$200 MBq to $\sim$378 MBq. These are then compared to the standard single-pixel (photoelectric) measurements. We find that the signal-to-random background ratios (SBRs) obtained from the polarization-correlated events for Compton scattering angles $θ_{1,2}\in[72^{\circ}, 90^{\circ}]$ and azimuthal angle difference $Δφ=90^{\circ}\pm20^{\circ}$ are consistently higher than those from single-pixel events, with the ratio of their SBR values of 1.23. The SBR of the selected events also increases with the polarimetric modulation factor $μ$, gaining $\sim$50\% in value during the experiment.

Figures (8)

• Figure 1: Simplified schematics of the experimental setup. The figure represents two segmented scintillating GAGG detector modules (in yellow) with SiPMs (in blue) in a coincidence measurement setup, read-out by the ToFPET2 ASIC system. The source (in green) emits two polarization-correlated annihilation photons (in purple) that undergo Compton scattering in their respective detector modules with Compton scattering angles $\theta_{1,2}$ and azimuthal scattering angles $\phi_{1,2}$.
• Figure 2: Acceptance corrected $\phi_1$-$\phi_2$ distribution at the sixth measurement position. The Compton scattering angle range was selected as $72^{\circ}<\theta<90^{\circ}$. The highlighted regions denote the azimuthal difference range of events selected for further analysis. The regions shaded in blue correspond to the $70^{\circ}<\Delta\phi<110^{\circ}$ , and denote the highest probability of polarization correlations, while the regions shaded in pink, with $-20^{\circ}<\Delta\phi<20^{\circ}$ , denote the lowest probability of two photons having correlated polarizations.
• Figure 3: Coincidence time spectra for a) single-pixel (photoelectric) events and b) polarization-correlated Compton events. The selected signal (in green) is estimated as $\pm$3$\sigma$ from the mean calculated by the Gaussian fit (red curve) to the highest peak in each spectrum. The random background (in blue) is estimated as 6$\sigma$ area, removed from the calculated mean by $\pm$15$\sigma$.
• Figure 4: The measured values of the polarimetric modulation factor $\mu$ exhibit higher values at lower source activity A levels for all selected event topologies.
• Figure 5: The variation of SBR of different polarization-correlated events' topologies compared to the single-pixel events with the source activity A.