Performance evaluation of a PET/MR scanner with multiple positron-­emitting radionuclides

Introduction: Although most PET/MR imaging studies are conducted using ¹⁸F, other positron-emitting radionuclides, such as ⁶⁴Cu, ⁶⁸Ga, ⁸⁹Zr and ¹²⁴I are also popular options, and the availability of them in clinical settings is boosting. These radionuclides, especially the non-pure positron emitters, differ from ¹⁸F in several aspects, including half life, positron energy, positron range, positron branching ratio and gamma emission profile. All these impact on the signal acquisition and imaging quality of the PET/MR scanner. In this study, we characterize the performance of the UIH PET/MR scanner with multiple positron-emitting radionuclides to verify the imaging ability of the system.

Methods: The PET/MR scanner was a United Imaging Healthcare uPMR 790 HD TOF PET/MR, with a field strength of 3.0 T. [¹⁸F]FDG was produced with the medical cyclotron, [⁶⁸Ga]GaCl₃ was obtained from a ⁶⁸Ge/⁶⁸Ga generator, while [⁸⁹Zr]Zr-oxalate and [¹²⁴I]NaI were manufactured through solid target irradiation and subsequent purification steps. Evaluation methods were developed according to the NEMA NU2-2012 standard for performance measurements of positron emission tomographs. In the original standard, the types of radionuclides being used to caliber the system were restricted, however, the application was expanded to includes all radionuclides involved for the purpose of this research. The IEC image quality (IQ) phantom was used to validate the imaging ability of the scanner on these radionuclides. The background activity in the phantom was adjusted to around 5.3 kBq/mL in all cases. Point sources were used to calibrate the spatial resolution of the system. The point sources were manufactured by introducing a tiny droplet of concentrated radioactive solution into the capillary tubes. A custom jig was designed to hold the sources to the position as specified by the NEMA standard.

Results: The reconstructed images of the IQ phantom were obtained for each radionuclide, and subsequent quantification were accordingly performed. The images for all radionuclides retrieves the SUV of lung insert, the cold spheres, the background, and the hot spheres accurately. Due to the robust reconstruction algorithm, prompt gamma emission were well accounted for and the image quality of non-pure gamma emitters ⁶⁸Ga and ¹²⁴I are also acceptable. The spatial resolution of these radionuclides was obtained by measuring the corresponding FHWMs of the point source response functions reconstructed by a filtered back projection algorithm. Conclusion: The PET/MR scanner evaluated showed satisfactory performance with ⁶⁸Ga, ⁸⁹Zr and ¹²⁴I. All radionuclides perform well in terms of the image quality and the spatial resolution. These findings lay a sound foundation for clinical image acquisition. Acknowledgements: The work is supported by Beijing Municipal Administration of Hospitals-Yangfan Project (ZYLX201816).