[Paper Review] Towards 2+1 photon tomography: Energy-based selection of two 511 keV photons and a prompt photon with the J-PET scanner

R. Masełek, W. Krzemień+32 more|arXiv (Cornell University)|Mar 2, 2018

Medical Imaging Techniques and ApplicationsMedicine1 references4 citations

TL;DR

This paper proposes an energy-based discrimination method in the J-PET scanner to distinguish 511 keV annihilation photons from prompt photons emitted during nuclear de-excitation in 2+1 photon emission tomography. By setting an optimal energy threshold of ~375 keV, the method achieves ~100% efficiency for 511 keV photons, ~75% purity, and ~70% efficiency and purity for prompt photons, with 44 Sc identified as the most suitable isotope for medical applications.

ABSTRACT

The possibility to separate signals caused by 511 keV photons created in annihilation of electron-positron pairs and the so-called prompt photons from nuclei de- excitation is investigated. It could potentially be used to improve the quality of reconstructed images in the J-PET scanner in 2+1 photon tomography. Firstly, a research is conducted for several radioisotopes that decay via \b{eta}+ decay followed by de-excitation of an excited nucleus. Efficiency, purity and false positive rate are calculated for each isotope as a function of energy deposited threshold, with a hypothesis that signals caused by 511 keV photons deposit smaller values of energy than 1 z 13the selected threshold, while prompt photons deposit larger energy than the threshold. Analysis of the results accompanied with physical properties of radioisotopes suggests using 44 Sc, which is the most promising candidate for medical applications. With the use of GATE and J-POS simulation software, in-phantom scattering was introduced and the best energy deposited threshold value was estimated to be approximately 375 keV. It corresponds to almost 100% efficiency for 511 keV signals, 75% purity for 511 keV photons, and approximately 70% efficiency and purity for prompt photons.

Motivation & Objective

• To improve image quality in 2+1 photon emission tomography by distinguishing prompt photons from 511 keV annihilation photons in the J-PET scanner.
• To evaluate the feasibility of energy-based discrimination between 511 keV photons and prompt photons from nuclear de-excitation.
• To identify the most suitable radioisotope for medical applications based on detection efficiency, purity, and false positive rate.
• To determine the optimal energy deposited threshold for maximizing discrimination performance in a realistic in-phantom scattering environment.
• To validate the method using GATE and J-POS simulations with realistic scattering effects included.

Proposed method

• Energy deposition thresholds were applied to differentiate 511 keV photons (lower energy deposition) from prompt photons (higher energy deposition).
• Simulations using GATE and J-POS software modeled photon interactions in a phantom, including scattering effects.
• For each radioisotope, efficiency, purity, and false positive rate were calculated as functions of the energy threshold.
• The physical properties of isotopes, including decay schemes and prompt photon energies, were analyzed to identify the most suitable candidate.
• The optimal threshold was determined by maximizing the trade-off between 511 keV photon detection efficiency and prompt photon purity.
• The analysis considered realistic conditions such as Compton scattering and detector response functions in a human-like phantom.

Experimental results

Research questions

• RQ1 Can energy-based discrimination effectively separate 511 keV annihilation photons from prompt photons in 2+1 photon emission tomography?
• RQ2 Which radioisotope from the studied set offers the best balance of detection efficiency, purity, and low false positive rate for medical imaging?
• RQ3 What is the optimal energy deposited threshold that maximizes the performance of the discrimination method under realistic in-phantom scattering conditions?
• RQ4 How do scattering effects in a phantom influence the energy spectrum and the effectiveness of energy-based selection?
• RQ5 To what extent does the method preserve signal integrity while rejecting background from prompt photons?

Key findings

• An energy deposited threshold of approximately 375 keV was identified as optimal for distinguishing 511 keV photons from prompt photons in the J-PET scanner.
• At this threshold, the method achieves nearly 100% efficiency for detecting 511 keV photons.
• The purity of 511 keV photon detection reaches approximately 75% at the optimal threshold.
• The method achieves around 70% efficiency and 70% purity for prompt photons, indicating strong discrimination capability.
• Among the studied isotopes, 44 Sc was identified as the most promising candidate for medical applications due to its favorable decay characteristics and discrimination performance.
• Simulations including in-phantom scattering confirmed that the selected threshold maintains high performance under realistic imaging conditions.