Comparison of quantitative image quality of cardiac computed tomography between raw-data-based and model-based iterative reconstruction algorithms with an emphasis on image sharpness

Background Image sharpness is commonly degraded on cardiac CT images reconstructed using iterative reconstruction algorithms. Objective To compare the image quality of cardiac CT between raw-data-based and model-based iterative reconstruction algorithms developed by the same CT vendor in children and young adults with congenital heart disease. Materials and methods In 29 patients with congenital heart disease, we reconstructed 39 cardiac CT datasets using raw-data-based and model-based iterative reconstruction algorithms. We performed quantitative analysis of image sharpness using distance(25-75%)and angle(25-75%)on a line density profile across an edge of the descending thoracic aorta in addition to CT attenuation, image noise, signal-to-noise ratio and contrast-to-noise ratio. We compared these quantitative image-quality measures between the two algorithms. Results CT attenuation did not show significant differences between the algorithms (P>0.05) except in the aorta. Image noise was small but significantly higher in the model-based algorithm than in the raw-data-based algorithm (4.8 +/- 2.3 Hounsfield units [HU] vs. 4.7 +/- 2.1 HU,P<0.014). Signal-to-noise ratio (110.2 +/- 50.9 vs. 108.4 +/- 50.4,P=0.050) and contrast-to-noise ratio (91.0 +/- 45.7 vs. 89.6 +/- 45.1,P=0.063) showed marginal significance between the two algorithms. The model-based algorithm showed a significantly smaller distance(25-75%)(1.4 +/- 0.4 mm vs. 1.6 +/- 0.3 mm,P<0.001) and a significantly higher angle(25-75%)(77.0 +/- 4.3 degrees vs. 74.1 +/- 5.7 degrees,P<0.001) than the raw-data-based algorithm. Conclusion Compared with the raw-data-based algorithm, the model-based iterative reconstruction algorithm demonstrated better image sharpness and higher image noise on cardiac CT in patients with congenital heart disease.