Application of Iterative Reconstruction to Reduce Noise in the Low-Dose Chest CT Imaging of Thoracic Outlet Level

赵永为,王鹤,王霄英,刘建新
DOI: https://doi.org/10.3969/j.issn.1000-0313.2013.03.016
2013-01-01
Abstract:Objective:To study the ability of inhibition of noise in the ultra-low-dose chest spiral CT scan of the thoracic outlet by using model-based iterative reconstruction(MBIR) technology.Methods:Twenty consecutive patients who underwent ultra-low-dose CT scan at the level of thoracic outlet were enrolled in this study.Scan range covered from bilateral carotid roots to the second thoracic vertebrae.Scanning conditions:120kVp,using tube current modulation technology,10~400mA,noise index 40,reconstruction thickness 0.625mm,slice distance 0.625mm.According to the reconstruction mode,the CT images were divided into ASIR group and MBIR group.Images at the same level of the two groups were taken at thoracic outlet,and regions of interest(ROI) were drawn in the same area in bilateral erector spinae muscles,thyroid,lung apex,vertebra,and tracheal lumen.CT value and SD value were measured on a AW4.5 workstation.CT value and SD value of the ROI in the 2 groups were compared by paired sample t-test respectively.There was significant difference when P0.05.Results:Average CT value of ROI of ASIR group was(349.86±96.70)HU,average SD value was(82.89±26.05),Meanwhile,those of MBIR group were(328.89±160.29)HU and(31.03±16.19),There was no significant difference between the CT values of the ROI of the 2 groups(P 0.05).There was significant difference between the SD values of the ROI of the 2 groups(P0.05).Conclusion:MBIR technology can effectively reduce the image noise of the thoracic outlet level compared to ASIR reconstruction images.
What problem does this paper attempt to address?