Adaptive Statistical Iterative Reconstruction 기법의 적용에 따른 Computed Tomography 영상의 화질과 선량 연구

A Study on Image Quality and Dose of CT Image Using Adaptive Statistical Iterative Reconstruction Algorithm

Purpose Recently many studies had been applied to minimize radiation exposure while also obtaining suitable image quality for diagnosis. This study was focused on the evaluation of image quality and dose of CT image which were applied with adaptive statistical iterative reconstruction(ASIR) to examine the usefulness of image diagnosis. Materials and methods CT scans were performed with a 64-slice MIDCT scanner(GE LightSpeed VCT XTe, waukesha, Wisconsin, USA). CTDI measurement was used for comparing doses while the ACR Phantom was used for image quality comparison. The ASIR was used for image reconstruction which create a noise map based on the filtered backprojection(FBP) and organized a statistical model and blends it with appropriately original FBP image. For image quality comparison in phantom study, CTDIVOL 14.76 mGy for non-ASIR was applied and evaluations were clone by analyzing items such as noise level, olv-contrast and high-contrast resolution. In patient study, test were clone by CTDIVOL 9.45 mGy (standard chest CT dose) and 0.90 mGy, 0.45 mGy, and 0.36 mGy(low dose CT) were delivered with standard and bone kernel. Results In phantom study showed that the upward values of spatial resolution were generally similar among kernels and low-resolution kernels showed excellence of contrast to noise(CNR) above 1.0 compared to high-resolution kernels in ASIR 50 and showed improved CNR even though decreased of dose 40% above. With same Parameters noise level(NL) showed 2% in FBP image, 1.5% in ASIR 30% above, and 1% or less in ASIR 100%. Differentiating by the naked eye was difficult, but the differences between each kernel of ASIR were recognized by measuring resolution and noise by quantitative analysis. The effective dose of the low-dose chest examination held as a patient study was only 0.3 mSv or less. It showed reduction of 90% compared to the standard dose, and similar image quality was maintained by applying ASIR 30~60%. However, high-resolution kernels With high ASIR showed a severe level of noise including streak artifacts. Conclusion CT image reorganization using ASIR is considered to be a useful dose reducing method while maintaining the image quality, when comparing with the existing filtered back-projection (FBP) reorganized images.