Linearity, Recovery Coefficient, Partial Volume Effect and Spill-Over Effect Evaluation for ${ }^{18}\mathrm{F}$ Imaging in the easyPET.3D

The easyPET.3D is a compact, cost-effective benchtop Positron Emission Tomography (PET) system, developed to address the limitations of conventional PET scanners, which are often expensive, complex, and inaccessible to many research institutions. Designed for small animal imaging, the easyPET.3D employs a dual-rotation acquisition mechanism that simplifies the system architecture while maintaining high image quality. This study aims to evaluate the system's reproducibility and performance in terms of key image quality parameters: uniformity, partial volume effect, and spillover effect. A NEMA NU-4 image quality phantom was filled with 6.5 MBq of fluorine-18 and scanned over a period of six hours. Static images were reconstructed using the List-Mode Maximum-Likelihood Expectation-Maximisation algorithm (40 iterations, voxel size: $\text{1. 0 0} \times \text{1. 0 0} \times \text{1. 1 4} \text{~ m m}^{\text{3}}$). The activity linear recovery was achieved for all tested activity concentrations, ranging up to $0.05\ \text{MBq} / \text{mL}$. Uniformity was stable across most conditions, with coefficient of variation (COV) values below 10 %, except at 1 MBq (15 %) due to limited counting statistics. Recovery coefficients were mainly reliable for rods $\geq 2 ~\text{mm}$ (COV: $4-10 \%$), while the smaller rod (1 mm) showed increased variability (COV: 14-35%). Spillover ratio values were higher than ideal, reflecting the absence of attenuation and scatter correction, and showed limited reproducibility (COV: 4-18%). The limitations in low-activity (3.7 MBq) and small-volume (1 mm) measurements observed highlight areas for future optimisation. Overall, the easyPET.3D provides consistent and reproducible measurements for the activity levels employed in preclinical studies, offering reliable performance for quantitative PET imaging in academic and research environments.