Influence of Ultra-Low-Dose CT on PET Image Quantification and Visual Assessment

Recent advances in high-sensitivity PET allow for significantly reduced radiation doses, making the CT radiation dose a larger fraction of the total scan dose. This is the first study to the best of our knowledge that evaluates the feasibility and efficacy of an ultra-low-dose CT (ULD-CT) protocol, incorporating a Sn filter, by reducing radiation exposure while maintaining PET image quality and quantitative accuracy in oncology patients. Methods: The study involved 29 oncology patients. FDG PET/CT scans were performed using both low-dose CT (LD-CT) and ULD-CT protocols. The ULD-CT protocol used the Sn filter with 100 kVp and 6 mAs, whereas the LD-CT protocol, optimized for PET attenuation correction and anatomic localization, was performed without the Sn filter, using 100-120 kVp and 12-30 mAs. Quantitative and qualitative comparisons of PET images were made, focusing on SUVs, Patlak analysis, and diagnostic quality on different tissues. Results: The ULD-CT protocol reduced the CT radiation dose by 97%, from 1.93 ± 0.61 mSv to 0.059 ± 0.026 mSv. SUV comparisons showed relative variations of less than 3%. The net rate of influx values resulting from the Patlak reconstruction showed negligible differences across various tissues between ULD-CT and LD-CT PET reconstructions. Image quality assessments revealed no significant differences between ULD-CT and LD-CT PET reconstructions, with consistently high scores (greater than 4 on a 5-point Likert scale). Lesion detection analysis showed strong intraobserver agreement (intraclass correlation coefficient > 0.85) without significant differences in lesion localization. Conclusion: The ULD-CT protocol substantially reduces radiation exposure in PET/CT while maintaining diagnostic quality of PET scans and introducing negligible quantitative effects, thereby paving the way for optimizing CT radiation protocols in PET/CT imaging.