Accuracy and radiation dose of CT-based attenuation correction for small animal PET: A Monte Carlo simulation study

Small animal PET allows qualitative assessment and quantitative measurement of biochemical processes in vivo, but the accuracy and reproducibility of imaging results can be affected by several parameters. The first aim of this study was to investigate the performance of different CT-based attenuation correction strategies and assess the resulting impact on PET images. The absorbed dose in different tissues caused by scanning procedures was also discussed to minimize biologic damage generated by radiation exposure due to PET/CT scanning. A small animal PET/CT system was modeled based on Monte Carlo simulation to generate imaging results and dose distribution. Three energy mapping methods, including the bilinear scaling method, the dual-energy method and the hybrid method which combines the kVp conversion and the dual-energy method, were investigated comparatively through assessing the accuracy of estimating linear attenuation coefficient at 511 keV and the bias introduced into PET quantification results due to CT-based attenuation correction. Our results showed that the hybrid method outperformed the bilinear scaling method, while the dual-energy method achieved the highest accuracy among the three energy mapping methods. Overall, the accuracy of PET quantification results have similar trend as that for the estimation of linear attenuation coefficients, whereas the differences between the three methods are more obvious in the estimation of linear attenuation coefficients than in the PET quantification results. With regards to radiation exposure from CT, the absorbed dose ranged between 7.29-45.58 mGy for 50-kVp scan and between 6.61-39.28 mGy for 80-kVp scan. For ¹⁸F radioactivity concentration of 1.86×10⁵ Bq/ml, the PET absorbed dose was around 24 cGy for tumor with a target-to-background ratio of 8. The radiation levels for CT scans are not lethal to the animal, but concurrent use of PET in longitudinal study can increase the risk of - iological effects. The hybrid method has been demonstrated to be an energy mapping method that could perform accurate conversion from CT numbers to linear attenuation coefficients at 511 keV without increasing CT radiation exposure, so it should be a proper method for longitudinal imaging studies performed on small animal PET/CT systems.