Prompt-gamma Energy Spectrum-based Simultaneous Dual-isotope PET Imaging

Yu Shi,Yirong Wang,Fanzhen Meng,Jianwei Zhou,Bo Wen,Fei Kang,Shouping Zhu
DOI: https://doi.org/10.1109/nss/mic44845.2022.10398938
2022-01-01
Abstract:Simultaneous dual-isotope PET imaging has great potential in clinical applications. The main challenge of dual-isotope PET imaging is that the coincidences from different nuclides cannot be distinguished by energy. According to the physical property of non-pure positron-emitting isotopes that an extra gamma photon is produced almost simultaneously with positron annihilation, the separation of the pure isotopes and the non-pure isotopes can be achieved based on the triple coincidence technique. However, not only do non-pure isotopes produce triple coincidences, but pure isotopes also generate random triple coincidences, which eventually lead to residues of the pure isotopes in the non-pure isotopes. In this study, a triple-coincidence correction strategy based on the prompt-gamma energy spectrum is proposed. First, the mixed energy spectrum of all the prompt photons in the triple coincidences is obtained. Then a Gaussian function is used to fit the energy spectrum of the ‘pseudo' prompt photons from the pure isotope, and thus the spectrum of the non-pure isotope is obtained from the mixed spectrum. According to both of the energy spectra, the number of triple coincidences generated by the non-pure isotope on each line of response is obtained by the moment estimation method. The phantom experiments of 18 F and 124 I show that the proposed method improves the accuracy of reconstructed images qualitatively and quantitatively, compared with other triple-coincidence methods.
What problem does this paper attempt to address?