Application of low-tube current using adaptive statistic iterative reconstruction in the accuracy of spinal quantitative CT

Objective: To investigate the reproducibility and accuracy of spinal BMD (bone mineral density) by low mA using adaptive statistic iterative reconstruction based on phantom model. Methods European spine phantom (ESP) was scanned with 5 different tube currents (40, 60, 80, 100 and 120 mA) on the GE Revolution prototype. All data were transferred to the quantitative CT (QCT) PRO workstation for measuring the bone mineral density (BMD). And all data were reconstructed using different adaptive statistical iterative reconstruction veo (ASiR-V) weighting percentages from ASiR-V 0 filtered back projection (FBP) to ASiR-V 100%, at interval of 20%. Quantitative measurements of CT value, noise, and contrast noise ratio (CNR) of L1, L2 and L3 were measured in each group. The single-factor analysis of variance (ANOVA) was used to compare the bone mineral density values of different mA and ASiR-V weighting percentages. Results: Volume CT dose index (CTDI_vol) and dose length product (DLP) were positively correlated with tube current (r = 1). The BMD of ESP phantoms had no statistically significant differences among the multi-center lumbar spines L1, L2 and L3 at different doses under the same ASiR-V weighting percentages (P > 0.05), as well as at the same dose under different ASiR-V weighting percentages (P > 0.05). The error of bone density accuracy was within 6%. Conclusions: Low dosage of 120 kV and 40 mA using ASiR-V without affecting the accuracy of BMD has the potential effect to reduce radiation dose without compromising image quality. (authors)