Imai Rumi, Miyazaki Osamu, Horiuchi Tetsuya, Asano Keisuke, Nishimura Gen, Sago Haruhiko, Nosaka Shunsuke
1 Department of Radiology, National Center for Child Health and Development, 2-10-1 Okura, Setagaya-ku, Tokyo, 157-8535 Japan.
2 Department of Radiology, Tokyo Metropolitan Children's Medical Center, Tokyo, Japan.
AJR Am J Roentgenol. 2017 Jun;208(6):1365-1372. doi: 10.2214/AJR.16.17593. Epub 2017 May 2.
Prenatal diagnosis of skeletal dysplasia by means of 3D skeletal CT examination is highly accurate. However, it carries a risk of fetal exposure to radiation. Model-based iterative reconstruction (MBIR) technology can reduce radiation exposure; however, to our knowledge, the lower limit of an optimal dose is currently unknown. The objectives of this study are to establish ultra-low-dose fetal CT as a method for prenatal diagnosis of skeletal dysplasia and to evaluate the appropriate radiation dose for ultra-low-dose fetal CT.
Relationships between tube current and image noise in adaptive statistical iterative reconstruction and MBIR were examined using a 32-cm CT dose index (CTDI) phantom. On the basis of the results of this examination and the recommended methods for the MBIR option and the known relationship between noise and tube current for filtered back projection, as represented by the expression SD = (milliamperes), the lower limit of the optimal dose in ultra-low-dose fetal CT with MBIR was set. The diagnostic power of the CT images obtained using the aforementioned scanning conditions was evaluated, and the radiation exposure associated with ultra-low-dose fetal CT was compared with that noted in previous reports.
Noise increased in nearly inverse proportion to the square root of the dose in adaptive statistical iterative reconstruction and in inverse proportion to the fourth root of the dose in MBIR. Ultra-low-dose fetal CT was found to have a volume CTDI of 0.5 mGy. Prenatal diagnosis was accurately performed on the basis of ultra-low-dose fetal CT images that were obtained using this protocol. The level of fetal exposure to radiation was 0.7 mSv.
The use of ultra-low-dose fetal CT with MBIR led to a substantial reduction in radiation exposure, compared with the CT imaging method currently used at our institution, but it still enabled diagnosis of skeletal dysplasia without reducing diagnostic power.
通过三维骨骼CT检查对骨骼发育异常进行产前诊断具有很高的准确性。然而,这会使胎儿面临辐射风险。基于模型的迭代重建(MBIR)技术可以减少辐射暴露;然而,据我们所知,目前最佳剂量的下限尚不清楚。本研究的目的是建立超低剂量胎儿CT作为骨骼发育异常产前诊断的方法,并评估超低剂量胎儿CT的合适辐射剂量。
使用32 cm CT剂量指数(CTDI)模体研究自适应统计迭代重建和MBIR中管电流与图像噪声之间的关系。根据该检查结果以及MBIR选项的推荐方法,以及滤波反投影中噪声与管电流的已知关系(用表达式SD =(毫安)表示),设定了MBIR超低剂量胎儿CT中最佳剂量的下限。评估使用上述扫描条件获得的CT图像的诊断能力,并将超低剂量胎儿CT的辐射暴露与先前报告中记录的辐射暴露进行比较。
在自适应统计迭代重建中,噪声几乎与剂量的平方根成反比增加,在MBIR中与剂量的四次方根成反比增加。发现超低剂量胎儿CT的容积CTDI为0.5 mGy。根据使用该方案获得的超低剂量胎儿CT图像准确地进行了产前诊断。胎儿的辐射暴露水平为0.7 mSv。
与我们机构目前使用的CT成像方法相比,使用带有MBIR的超低剂量胎儿CT可大幅减少辐射暴露,但仍能在不降低诊断能力的情况下诊断骨骼发育异常。