²²³Ra SPECT/CT (Single Photon Emission Computed Tomography/Computed Tomography) imaging provides critical information for evaluating the efficacy of bone-targeted radiotherapy in patients with metastatic castration-resistant prostate cancer (mCRPC).

This study aims to explore the optimal imaging parameters for ²²³Ra SPECT/CT using three standard phantoms: the Carlson model, PET/CT model, and Hoffman 3D brain model.

Firstly, quality control tests were performed on all three phantoms using a GE NM/CT 670 ES system with ⁹⁹Tc^m to establish standard reference images. Then, ²²³Ra SPECT/CT imaging was conducted by injecting 3.7 MBq of ²²³RaCl₂ into each phantom, with acquisition parameters set at 84 keV and 140 keV energy peaks with 10% windows, using both LEHP and HEGP collimators, and acquisition times ranging from 30 min to 100 min. Finally, image reconstruction was performed using Butterworth filters (f_c=0.48, p=10) and three-dimensional ordered subset expectation maximization (3D OSEM) algorithms with 5 subsets and 10 iterations, followed by qualitative assessment of image quality by two experienced nuclear medicine physicians.

The ²²³Ra SPECT/CT images acquired with HEGP collimators demonstrate significantly better resolution and contrast compared to those with LEHP collimators. In the Carlson model, the minimum detectable hot and cold zone diameters are both 22.3 mm, with acceptable linearity but suboptimal uniformity. PET/CT model ²²³Ra SPECT/CT images are acceptable, but there are certain differences between the transverse, coronal, and sagittal image fusion matching degree and attenuation correction dispersion index compared to the ⁹⁹Tc^m SPECT/CT quality control images. The Hoffman brain phantom ²²³Ra SPECT/CT images show less distinct brain sulci, gyri, and basal ganglia nuclear groups, indicating a need to further improve image quality.

²²³Ra SPECT/CT imaging is feasible for clinical applications, with optimal parameters including an 84 keV±20% energy window and HEGP collimation. While the current protocol achieves diagnostic-quality images with a spatial resolution of 22.3 mm for both hot and cold lesions, further parameter optimization is needed to improve image uniformity and detailed visualization of complex structures. These findings provide a foundation for clinical ²²³Ra SPECT/CT imaging protocols that can enable more accurate assessment of ²²³Ra therapy distribution and response.