Department of Physics, Faculty of Sciences and Mathematics, Diponegoro University, Jl. Prof. Soedarto SH, Tembalang, Semarang 50275, Central Java, Indonesia.
Department of Physics, Faculty of Mathematics and Natural Sciences, Bandung Institute of Technology, Bandung, West Java, Indonesia.
Biomed Phys Eng Express. 2020 Sep 29;6(6). doi: 10.1088/2057-1976/abb2f8.
The purpose of this study was to develop a computational phantom for validation of automatic noise calculations applied to all parts of the body, to investigate kernel size in determining noise, and to validate the accuracy of automatic noise calculation for several noise levels. The phantom consisted of objects with a very wide range of HU values, from -1000 to +950. The incremental value for each object was 10 HU. Each object had a size of 15 × 15 pixels separated by a distance of 5 pixels. There was no dominant homogeneous part in the phantom. The image of the phantom was then degraded to mimic the real image quality of CT by convolving it with a point spread function (PSF) and by addition of Gaussian noise. The magnitude of the Gaussian noises was varied (5, 10, 25, 50, 75 and 100 HUs), and they were considered as the ground truth noise (N). We also used a computational phantom with added actual noise from a CT scanner. The phantom was used to validate the automated noise measurement based on the average of the ten smallest standard deviations (SD) from the standard deviation map (SDM). Kernel sizes from 3 × 3 up to 27 × 27 pixels were examined in this study. A computational phantom for automated noise calculations validation has been successfully developed. It was found that the measured noise (N) was influenced by the kernel size. For kernels of 15 × 15 pixels or smaller, the Nvalue was much smaller than the N. For kernel sizes from 17 × 17 to 21 × 21 pixels, the Nvalue was about 90% of N. And for kernel sizes of 23 × 23 pixels and above, Nis greater than N. It was also found that even with small kernel sizes the relationship between Nand Nis linear with Rmore than 0.995. Thus accurate noise levels can be automatically obtained even with small kernel sizes without any concern regarding the inhomogeneity of the object.
这项研究的目的是开发一种计算体模，用于验证应用于全身各部位的自动噪声计算的准确性，研究核大小对噪声的影响，并验证自动噪声计算在多个噪声水平下的准确性。体模由具有非常宽 HU 值范围（从-1000 到+950）的物体组成。每个物体的增量值为 10 HU。每个物体的大小为 15×15 像素，间隔 5 像素。体模中没有主导的均匀部分。然后通过卷积点扩散函数（PSF）和添加高斯噪声来使体模图像降级，以模拟 CT 的真实图像质量。高斯噪声的幅度变化（5、10、25、50、75 和 100 HU），并将其视为实际噪声（N）。我们还使用了一个带有 CT 扫描仪添加实际噪声的计算体模。该体模用于验证基于标准偏差图（SDM）中十个最小标准偏差（SD）的平均值的自动噪声测量。在这项研究中，检查了从 3×3 到 27×27 像素的核大小。已经成功开发了一种用于自动噪声计算验证的计算体模。研究发现，测量噪声（N）受核大小的影响。对于 15×15 像素或更小的核，N 值比 N 小得多。对于 17×17 到 21×21 像素的核大小，N 值约为 N 的 90%。对于 23×23 像素及以上的核大小，N 大于 N。还发现，即使核尺寸较小，N 与 N 之间的关系也是线性的，R 值大于 0.995。因此，即使使用较小的核尺寸，也可以自动获得准确的噪声水平，而无需担心物体的不均匀性。
