Simulations Show Quadra PET Scanner Boosts Positronium Imaging Sensitivity

Researchers published a simulation-based feasibility study in EJNMMI Physics demonstrating that the Biograph Vision Quadra positron emission tomography scanner can achieve significantly higher sensitivity for positronium lifetime imaging than previous prototype systems. By modeling the physics of photon interactions with the GATE software, the team estimated that the Quadra system could improve imaging sensitivity by a factor of 150 compared to the modular J-PET prototype used for the first in-vivo positronium lifetime images.

Positronium lifetime imaging represents a novel diagnostic approach that seeks to extract complementary information from PET scans beyond standard metabolic uptake. In conventional PET, the Standardized Uptake Value quantifies the concentration of a radiotracer in tissues, helping to identify regions of pathological metabolism. However, current systems do not exploit the physical properties of positronium, a metastable atom formed when a positron emitted by a radionuclide binds with an electron in the surrounding medium. The lifetime of this atom, particularly in its ortho-positronium state, is sensitive to the local nanostructure of the tissue and the concentration of bioactive molecules such as oxygen.

The core principle of positronium lifetime imaging relies on the detection of a specific triple coincidence event. When a radionuclide such as scandium-44 undergoes beta-plus decay, it emits a positron and a prompt gamma ray from the deexcitation of the daughter nucleus. The positron may form positronium, which eventually annihilates into two 511 keV photons. By registering the prompt gamma as a timestamp for positronium formation and the annihilation photons as the endpoint, the system can estimate the positronium lifetime. This measurement provides a qualitatively different contrast mechanism from standard SUV, potentially offering insights into intramolecular architecture that metabolic imaging cannot provide.

The study focused on the Biograph Vision Quadra, a crystal-based PET scanner with a 106 cm axial field-of-view, and compared its performance to the modular J-PET prototype and the Total Body J-PET scanner. The Quadra utilizes lutetium oxyorthosilicate scintillator crystals, which offer a higher probability of photon interaction compared to the plastic scintillators used in the J-PET systems. The simulation modeled the decay of four radioisotopes: gallium-68, scandium-44, sodium-22, and iodine-124. These isotopes were selected based on their current clinical use, feasibility for positronium imaging, and potential applications in thyroid cancer diagnosis and basic research.

A critical aspect of the simulation was the optimization of energy windows to distinguish between annihilation photons and prompt gamma rays. In standard PET, an energy window centered around 511 keV filters out scattered photons. For positronium lifetime imaging, a second energy window is required to register the prompt gamma, which typically has a higher energy than the annihilation photons. The researchers determined that the highest sensitivity is achieved when the energy window for the prompt gamma is placed immediately adjacent to the upper limit of the annihilation photon window. For most isotopes, this configuration maximizes the detection efficiency while maintaining high purity of the triple coincidence events.

The results indicated that the organ-wise sensitivity of the Quadra for positronium lifetime imaging varies by isotope. When background radiation and hardware limitations were included in the simulation, the estimated sensitivities were 9.22 cps/kBq for scandium-44, 10.46 cps/kBq for gallium-68, 5.91 cps/kBq for sodium-22, and 15.39 cps/kBq for iodine-124. The iodine-124 isotope performed best due to its prompt gamma energy of 603 keV, which allows for a more distinct separation from the annihilation photons despite some spectral overlap. The study noted that the introduction of background radiation and energy measurement limits reduced the ideal sensitivity by a factor of two to three, highlighting the importance of realistic modeling in feasibility studies.

Comparisons with the J-PET scanners revealed substantial improvements in sensitivity. The modular J-PET prototype, which has a 50 cm axial field-of-view and uses plastic scintillators, achieved a sensitivity of 0.062 cps/kBq for scandium-44. In contrast, the simulated Total Body J-PET scanner, with a 243 cm axial field-of-view, was estimated to achieve a sensitivity of 1.714 cps/kBq. The Quadra’s sensitivity for scandium-44 was found to be 150 times higher than that of the modular J-PET, while the Total Body J-PET was expected to be 30 times higher than the modular version.

The study also examined the impact of background radiation, including intrinsic radiation from the lutetium in the Quadra’s crystals and cosmic rays. The researchers added background data gathered from an empty Quadra scanner in Bern, Switzerland, to the simulation to account for false coincidences. This background was found to affect the sensitivity but did not alter the relative performance ranking of the different isotopes. The purity of the triple coincidence events remained high, with the probability of selecting true candidate triples exceeding 97% in most configurations. The accidental coincidence rate was kept low by simulating a low activity level of 1 MBq, which is typical for such feasibility studies.

For the iodine-124 isotope, the overlapping photopeaks of the annihilation and prompt gamma rays required a different optimization strategy. The researchers adjusted the upper limit of the lower energy window to 550 keV, which improved the figure of merit by 20% compared to the standard window ending at 585 keV. This adjustment allowed for a better balance between the purity of the prompt gamma selection and the relative efficiency of its detection. The study emphasizes that while higher statistics generally improve image quality, further investigations involving image reconstruction are necessary to fully assess the impact of these sensitivity gains on the final diagnostic images.

The findings suggest that long axial field-of-view PET scanners are well-suited for positronium lifetime imaging, addressing the low statistics issue that has limited previous in-vivo demonstrations. The Quadra’s ability to operate in singles mode acquisition, which allows for the registration of all photon interactions, is a key enabler for this technique. The study concludes that the whole-body sensitivity for positronium lifetime imaging will be more than one order of magnitude higher for the Total Body J-PET and two orders of magnitude higher for the Biograph Vision Quadra compared to the modular J-PET prototype.

The simulation study utilized the GATE software, version 9.0, an open-source Monte-Carlo tool developed by the OpenGATE Collaboration for medical imaging studies. The software was configured to model the radioactive beta-plus decays and the subsequent deexcitation of daughter nuclei, tracing the interaction of their products within the tomographs. A specific limitation of the software is that it does not simulate the interactions of positronium atoms with the surrounding medium, such as conversion or pick-off processes that typically shorten the ortho-positronium lifetime in biological tissues. Consequently, the researchers simulated the para-positronium decay into two photons and the emission of a prompt gamma with an uncorrelated direction relative to the annihilation photons, assuming the default vacuum lifetime. This approach allowed for the determination of system sensitivity, which is independent of the actual positronium lifetime in the tissue.

Event selection in the simulation required a post-processing step because the GATE software does not natively support the creation of triple coincidences involving two annihilation photons and one deexcitation photon. The analysis processed a series of singles events sorted in time, imposing a time window to identify candidate triple coincidences. A candidate was exported if exactly two interactions fell into the lower energy window and one into the higher energy window within the specified time frame. The time windows used were 4.7 nanoseconds for the Quadra and 3 nanoseconds for the J-PET systems, dictated by their respective geometrical acceptance. In experimental positronium lifetime imaging, these windows would need to be extended, potentially to 10 nanoseconds, to account for the mean positronium lifetime in various tissues, which ranges from approximately 1.4 to 4 nanoseconds.

The organ-wise sensitivity was calculated according to NEMA standards, defined as the ratio of registered true triple coincidences originating from within a specific slice to the activity present in that slice. The simulation employed a 183 cm linear source located at the central axis of the tomograph with a total activity of 1 MBq. The organ-wise sensitivity itself was defined as the mean of the sensitivities of slices within a ±4.5 cm region around the scanner’s center. This methodology allowed for a standardized comparison across the different scanner geometries and radioisotopes, ensuring that the reported sensitivities in counts per second per kilobecquerel were directly comparable.

The study highlights that the requirement for three-photon coincidence significantly reduces sensitivity compared to standard metabolic PET imaging. However, the trend toward constructing long axial field-of-view PET scanners offers a potential resolution to this challenge. The Biograph Vision Quadra, the PennPET Explorer, and the J-PET systems are among the scanners capable of singles mode acquisition, which is necessary for registering all photon interactions required for positronium lifetime imaging. The simulation results indicate that the determined sensitivities for the Quadra and Total Body J-PET are in the order of sensitivities of metabolic PET imaging with short axial field-of-view scanners based on pure positron emitter isotopes. This suggests that the technical hurdles of low statistics in positronium lifetime imaging may be mitigated by the increased geometric acceptance of modern total-body scanners.

Subject of Research: Feasibility simulation of positronium lifetime imaging with Biograph Vision Quadra and J-PET tomographs

Article Title: Feasibility simulation study of positronium lifetime imaging with the Biograph Vision Quadra and J-PET tomographs

Article References: Parzych, S., Niedźwiecki, S., Beyene, E. Y., Chug, N., Conti, M., Curceanu, C., Czerwiński, E., Das, M., Valsan Eliyan, K., Hajduga, J., Jalali, S., Kacprzak, K., Kaplanoglu, T., Kapłon, Ł., Kasperska, K., Khreptak, A., Korcyl, G., Kozik, T., Kumar, D., … Moskal, P. (2026). Feasibility simulation study of positronium lifetime imaging with the Biograph Vision Quadra and J-PET tomographs. EJNMMI Physics, 13(1), Article 102. https://doi.org/10.1186/s40658-026-00920-1

Image Credits: AI Generated

DOI: 10.1186/s40658-026-00920-1

Keywords: positronium lifetime imaging, PET scanner, Biograph Vision Quadra, J-PET, medical imaging, simulation study, radioisotopes, sensitivity, EJNMMI Physics, Feasibility, simulation, positronium