Component-Based Normalization for a Multi-Resolution DOI PET System

A component-based normalization method was developed for a preclinical PET protype system (SPET) which employs multi-resolution DOI detectors. The detector consists of dual-layer pixelated scintillators, with LYSO crystals measuring $0.86 \times 0.86 \times 5 \mathrm{~mm}^{3}$ arranged in a $30 \times 60$ array in the top layer and BGO crystals measuring $1.29 \times 1.29 \times 7.5 \mathrm{~mm}^{3}$ arranged in a $20 \times 40$ array in the bottom layer. Compared to direct normalization, the proposed component-based normalization method addresses challenges in data acquisition and storage efficiency by transforming the original PET system into a configuration with uniform-sized pseudo crystals.Two simulated scans were performed using cylindrical and annular sources to calculate four normalization component factors. The axial block profile factor bax were obtained by considering only the direct sinogram planes, while the axial geometric factor gax were computed for all combinations. The intrinsic efficiency factor $\varepsilon u($ were computed using a 3D fan-sum algorithm, and the transverse factor frdtr were computed by considering the transverse differences in the direct sinogram planes. A simulation using a large uniform cylinder, along with experimental data acquired using the NEMA NU-4 image quality phantom (IQP), were used to evaluate the performance of the algorithm in improving image uniformity and reducing artifacts. Through normalization, uniform reconstructed images were obtained for both simulated cylinder data and experimental IQP data.In conclusion, the proposed component-based normalization method greatly contributed to the production of high-quality images. By effectively addressing challenges in a multi-resolution DOI PET system, this approach leads to reliable reconstructed images from both simulated and experimental data. Future reserch will focus on using experimental data to improve the accuracy of normalization and to optimize the performance of the small animal PET system in biomedical research and preclinical studies.