Söderberg Marcus, Gunnarsson Mikael, Nilsson Mats
Department of Medical Radiation Physics, Malmö, Lund University, Skåne University Hospital, SE-205 02 Malmö, Sweden.
Radiat Prot Dosimetry. 2010 Apr-May;139(1-3):71-7. doi: 10.1093/rpd/ncq027. Epub 2010 Feb 19.
The purpose of this study was to verify and validate a noise simulation tool called Dose Tutor (VAMP GmbH) in terms of level and texture of the simulated noise. By adding artificial noise to measured computed tomography (CT) raw data, a scan acquired with a lower dose (mAs) than the actual one can be simulated. A homogeneous polyethylene phantom and an anthropomorphic chest phantom were scanned for different mAs levels, tube voltages, slice thicknesses and reconstruction kernels. The simulated noise levels were compared with the noise levels in real transverse slice images actually acquired with corresponding mAs values. In general, the noise comparisons showed acceptable agreement in magnitude (<20 % deviation in pixel standard deviation). Also, the calculated noise power spectra were similar, which indicates that the noise texture is correctly reproduced. In conclusion, this study establishes that the Dose Tutor might be a useful tool for estimating the dose reduction potential for CT protocols.
本研究的目的是在模拟噪声的水平和纹理方面验证和确认一种名为剂量导师（VAMP GmbH）的噪声模拟工具。通过向测量的计算机断层扫描（CT）原始数据中添加人工噪声，可以模拟出比实际剂量（毫安秒）更低的扫描。对均匀聚乙烯体模和拟人化胸部体模在不同的毫安秒水平、管电压、层厚和重建核下进行扫描。将模拟的噪声水平与实际以相应毫安秒值采集的真实横向切片图像中的噪声水平进行比较。总体而言，噪声比较在幅度上显示出可接受的一致性（像素标准差偏差<20%）。此外，计算出的噪声功率谱相似，这表明噪声纹理得到了正确再现。总之，本研究表明剂量导师可能是一种用于估计CT扫描方案剂量降低潜力的有用工具。
