Comparative Assessment of the Influences of Iterative Model Reconstruction (IMR) and Hybrid-Iterative Reconstruction (HIR) on Image Quality of 256-Slice Coronary Computed Tomography Angiography

Abstract

Background: The aim of this study was to evaluate the influences of iterative model reconstruction (IMR) and hybrid-iterative reconstruction (HIR) techniques on quantitative and qualitative image analysis of 256-slice coronary computed tomography angiography (CCTA). Methods: Sixty-one patients (30 men and 31 women with mean age of 60.68 ± 9.13 years) who had undergone CCTA by the 256-slice CT scanner were evaluated. The raw data were reconstructed using HIR and IMR algorithms. For objective assessment of image quality, parameters of noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were obtained for both reconstruction algorithms. For subjective assessment, two physician specialists evaluated image quality using a 5-point scale. Findings: The mean image noise on HIR and IMR images was 32.74 ± 6.04 and 25.15 ± 3.59, respectively. In IMR, the CNR method (28.81 ± 5.13) was significantly better than HIR method (23.06 ± 5.03) (P < 0.001). However, the HIR method (4.62 ± 0.39) was higher than the IMR method (4.48 ± 0.45) in terms of qualitative or subjective criteria (P < 0.001). Conclusion: The IMR method can improve quantitative criteria of image, reduce noise, and increase SNR and CNR of images better than HIR method. However, the images which reconstructed by HIR were superior to IMR in terms of qualitative criteria.

Keywords

• coronary vessels
• computed tomography angiography
• image reconstruction, image quality enhancement