Impact of different post-processing correction techniques on the quantitative results of ⁹⁹Tc^m SPECT/CT

Objective: To evaluate the effects of different sphere volumes, target background ratio (T/B) and post-processing correction techniques on the quantitative results of ⁹⁹Tc^m SPECT/CT. Methods: Six spheres with different diameters (37, 28, 22, 17, 13, 10 mm) in National Electrical Manufacturers Association International Electrotechnical Commission (NEMA IEC) models were filled with a mixture of 0.54 MBq/ml ⁹⁹Tc^m and iodixanol. The mixture iodine content was about 0.3% (135 mg), which led to different T/B (32:1, 16:1, 8:1, 4:1) by changing the radioactivity concentration of the cylinder. Routine imaging was performed on different T/B phantoms which were scanned by SPECT/CT. The CT threshold method was used for the delineation of volume of interest (VOI). Then the same processing correction technique and ordered-subsets expectation maximization (OSEM) parameters were used to calculate the radioactivity concentrations of different spheres, and further compared with the true values, and the accuracies were calculated. Pearson correlation analysis was applied to evaluate the relationships between sphere volume, T/B and quantitative results. The sphere with T/B of 32:1 and diameter of 37 mm were processed by 3 correction techniques (CT attenuation correction (CTAC)+scatter correction (SC)+resolution recovery (RR); CTAC+SC; CTAC+RR). One-way analysis of variance and the least significant difference t test were used to analyzed the effects of 3 correction techniques on the quantitative results and image contrasts. Results: There were significant relationships between the sphere volumes, T/B and the quantitative accuracy (r values: 0.757, 0.409, both P < 0.05). There were significant differences of 3 correction techniques on the quantitative results and image contrast (F values: 139.665 and 38.905, both P < 0.001). Among them, the quantitative error of CTAC+SC+RR was lower than that of CTAC+SC ((9.63 ± 8.82)% vs (38.89 ± 2.17)% P < 0.001), and similar to that of CTAC+RR ((8.70 ± 6.64)% P > 0.05). The quantitative error of CTAC+RR was lower than that of CTAC+SC (P < 0.001). The image contrast of CTAC+SC+RR was higher than that of CTAC+SC ((93.45 ± 0.91)% vs (92.41 ± 0.25)% P < 0.001) and the image contrast of CTAC+SC was higher than that of CTAC+RR ((91.37 ± 0.87)% P < 0.001). Conclusions: The larger sphere volume and the higher T/B, the more quantitative accuracy. The volume has a more significant effect on quantitative accuracy than T/B. Choosing the appropriate correction technique is helpful to quantitative accuracy improvement. It is suggested to use CTAC+SC+RR in quantitative processing. (authors)