CT coronary calcium scoring with tin filtration using iterative beam-hardening calcium correction reconstruction

Objectives To investigate the diagnostic accuracy of CT coronary artery calcium scoring (CACS) with tin pre-filtration (Sn100 kVp) using iterative beam-hardening correction (IBHC) calcium material reconstruction compared to the standard 120 kVp acquisition. Background Third generation dual-source CT (DSCT) CACS with Sn100 kVp acquisition allows significant dose reduction. However, the Sn100 kVp spectrum is harder with lower contrast compared to 120kVp, resulting in lower calcium score values. Sn100 kVp spectral correction using IBHC-based calcium material reconstruction may restore comparable calcium values. Methods Image data of 62 patients (56% male, age 63.9 ± 9.2years) who underwent a clinically-indicated CACS acquisition using the standard 120 kVp protocol and an additional Sn100 kVp CACS scan as part of a research study were retrospectively analyzed. Datasets of the Sn100 kVp scans were reconstructed using a dedicated spectral IBHC CACS reconstruction to restore the spectral response of 120 kVp spectra. Agatston scores were derived from 120 kVp and IBHC reconstructed Sn100 kVp studies. Pearson's correlation coefficient was assessed and Agatston score categories and percentile-based risk categorization were compared. Results Median Agatston scores derived from IBHC Sn100 kVp scans and 120 kVp acquisition were 31.7 and 34.1, respectively (p = 0.057). Pearson's correlation coefficient showed excellent correlation between the acquisitions (r = 0.99, p < 0.0001). Agatston score categories and percentile-based cardiac risk categories showed excellent agreement (ĸ = 1.00 and ĸ = 0.99), resulting in a low cardiac risk reclassification of 1.6% with the use of IBHC CACS reconstruction. Image noise was 24.9 ± 3.6HU in IBHC Sn100 kVp and 17.1 ± 3.9HU in 120 kVp scans (p < 0.0001). The dose-length-product was 13.2 ± 3.4 mGy cm with IBHC Sn100 kVp and 59.1 ± 22.9 mGy cm with 120 kVp scans (p < 0.0001), resulting in a significantly lower effective radiation dose (0.19 ± 0.07 mSv vs. 0.83 ± 0.33 mSv, p < 0.0001) for IBHC Sn100 kVp scans. Conclusion Low voltage CACS with tin filtration using a dedicated IBHC CACS material reconstruction algorithm shows excellent correlation and agreement with the standard 120 kVp acquisition regarding Agatston score and cardiac risk categorization, while radiation dose is significantly reduced by 75% to the level of a chest x-ray