ISSN: 2168-9784
Masafumi Uchida
Objectives: To evaluate the quality of three-dimensional (3D) CT angiography images of the liver using iterative model reconstruction (IMR).
Materials and Methods: Twenty-four patients with suspected disease of the pancreatobiliary system who had undergone CT reconstruction with filtered back projection (FBP) (Protocol A), iDose (hybrid IR) (Protocol B), and IMR (Protocol C) were evaluated. Contrast-to-noise ratios (CNR) relative to the liver, aorta, and portal vein were measured and recorded. The quality of 3D images under the same CT-threshold was graded on a scale of 1 (poor) to 5 (excellent).
Results: The CNR of the artery and portal vein was highest when using protocol C. Mean value (±standard deviation) of the artery-to-liver CNRs for each protocol was as follows: A (6.95 ± 1.53), B (12.05 ± 2.38), and C (33.12 ± 7.93). Mean value (±standard deviation) of the portal vein-to-liver CNRs for each protocol was as follows: A (2.81 ± 0.62), B (5.17 ± 1.40), and C (10.27 ± 3.99). The quality of 3D images significantly differed between the three methods, with that of Protocol C being the highest.
Conclusions: Compared to FBP and hybrid IR, abdominal CT scans via IMR produced higher-quality 3D
angiography images of the liver at the same radiation dose and with significant improvements in CNR.