Coronary CT angiography: Comparison of a novel iterative reconstruction with filtered back projection for reconstruction of low-dose CT—Initial experience

Objective: To prospectively compare subjective and objective image quality in 20% tube current coronary CT angiography (cCTA) datasets between an iterative reconstruction algorithm (SAFIRE) and traditional filtered back projection (FBP). Materials and methods: Twenty patients underwent a prospectively ECG-triggered dual-step cCTA protocol using 2nd generation dual-source CT (DSCT). CT raw data was reconstructed using standard FBP at full-dose (Group1a) and 80% tube current reduced low-dose (Group1b). The low-dose raw data was additionally reconstructed using iterative raw data reconstruction (Group2). Attenuation and image noise were measured in three regions of interest and signal-to-noise-ratio (SNR) as well as contrast-to-noise-ratio (CNR) was calculated. Subjective diagnostic image quality was evaluated using a 4-point Likert scale. Results: Mean image noise of group2 was lowered by 22% on average when compared to group1b (p < 0.0001–0.0033), while there were no significant differences in mean attenuation within the same anatomical regions. The lower image noise resulted in significantly higher SNR and CNR ratios in group2 compared to group1b (p < 0.0001–0.0232). Subjective image quality of group2 (1.88 ± 0.63) was also rated significantly higher when compared to group1b (1.58 ± 0.63, p = 0.004). Conclusions: Image quality of 80% tube current reduced iteratively reconstructed cCTA raw data is significantly improved when compared to standard FBP and consequently may improve the diagnostic accuracy of cCTA