OBJECTIVE: The spatial resolution of PET imaging is intrinsically limited by the finite range of positrons before annihilation, an effect that becomes increasingly relevant for high-energy emitters and for preclinical studies targeting small structures. In integrated PET/MRI systems, the magnetic field can modify positron trajectories, reducing their transverse spread and introducing anisotropic resolution effects. However, experimental evidence of the combined influence of positron energy and material density under high-field preclinical conditions remains limited. This work experimentally investigates the effect of a 9.4 T magnetic field on positron-range blurring for different radionuclides and tissue-equivalent phantoms in a preclinical PET/MRI environment. APPROACH: Experiments were performed using a high-resolution preclinical PET insert operating simultaneously inside a preclinical 9.4 T MRI system to study three positron-emitting radionuclides (¹⁸F, ⁸⁹Zr and ⁶⁸Ga). Capillary line sources were embedded in three tissue-equivalent phantoms with increasing density. Spatial resolution was quantified along the three spatial directions using the full width at half maximum (FWHM) and full width at tenth maximum (FWTM). Complementary measurements with a microDerenzo phantom were performed to assess rod resolvability. MAIN RESULTS: In low-density material, spatial resolution remains essentially unchanged by the magnetic field for all radionuclides. In contrast, in medium- and high-density materials, a marked transverse confinement is observed at 9.4 T for ⁶⁸Ga, with transverse FWTM values reduced by about 60% compared with 0 T, while the axial component remains largely unaffected. For ⁸⁹Zr, transverse improvements of about 25% are observed. These trends are consistent with the microDerenzo results: for ⁶⁸Ga, no rod sector is resolvable at 0 T, whereas at 9.4 T rods close to 1.0 mm become resolvable. SIGNIFICANCE: This study provides a systematic experimental assessment of positron-range confinement in high-field preclinical PET/MRI and demonstrates that the magnitude of magnetic-field-induced resolution improvements depends on both positron energy and material density.
Lopez-Berenguer et al. (Wed,) studied this question.