Impact of flat panel-imager veiling glare on scatter-estimation accuracy and image quality of a commercial on-board cone-beam CT imaging system
Impact of flat panel-imager veiling glare on scatter-estimation accuracy and image quality of a commercial on-board cone-beam CT imaging system
复制标题
DOI：
10.1118/1.4747260
复制
发表时间：
2012-09-01
期刊：
影响因子：
3.8
通讯作者：
Williamson, Jeffrey F.
中科院分区：
文献类型：
--
作者：
Lazos, Dimitrios;Williamson, Jeffrey F.
目的：本研究的目的是测量调制传递函数（MTF）的低频降（LFD），并研究其对锥束CT（CBCT）图像质量和散射测量精度的影响。方法：两种不同的实验方法被用来表征LFD和它的相关PSF的瓦里安OBI平板探测器系统：边缘响应函数（ERF）的方法和磁盘传递函数（DTF）的方法。通过将参数模型拟合到这些测量值的四个施加电压值（kVp）来估计PSF。所得到的PSF用于证明LFD对从合成数据集重建的CBCT图像中的图像对比度和CT数量准确性的影响，以及使用射束停止方法的散射测量的准确性。从测量的电流变液数据导出的MTF显示LFD从8%变化到10%。（在60 kVp）至10.5%（在120 kVp），而发现长PSF尾部的强度随着kVp的增加而增加。由ERF和DTF方法得到的杂散光线扩散函数具有很好的一致性。未校正的遮盖眩光降低了CBCT重建中的对比度和图像强度，靠近体模周边（在直径20 cm的水体模中为67个Hounsfield单位）和（在较小程度上）靠近不均匀性。蝴蝶结滤波器的使用减轻了这些影响。面纱眩光也导致约10%-15%的高估的散射-初级比测量时，与光束停止或光束停止array method.Conclusions：探测器PSF的长尾被发现有一个适度的依赖光束光谱，这是反映在MTF曲线LFD。我们的研究结果表明，未校正的面纱眩光会影响CBCT成像的定量准确性和对比度，基于平板成像仪。此外，它会导致高估的散射与初级比，测量与光束停止的方法。(c)2012年美国医学物理学家协会。[http：//dx.doi.org/10.1118/1.4747260]
Purpose: The purposes of this study is to measure the low frequency drop (LFD) of the modulation transfer function (MTF), associated with the long tails of the detector point spread function (PSF) of an on-board flat panel imager and study its impact on cone-beam CT (CBCT) image quality and scatter measurement accuracy.Methods: Two different experimental methods were used to characterize LFD and its associated PSF of a Varian OBI flat-panel detector system: the edge response function (ERF) method and the disk transfer function (DTF) method. PSF was estimated by fitting parametric models to these measurements for four values of the applied voltage (kVp). The resultant PSF was used to demonstrate the effect of LFD on image contrast and CT number accuracy in CBCT images reconstructed from synthetic datasets, as well as, accuracy of scatter measurements with the beam-stop method.Results: The MTFs derived from the measured ERF data revealed LFDs varying from 8% (at 60 kVp) to 10.5% (at 120 kVp), while the intensity of the long PSF tails was found to increase with increasing kVp. The veiling glare line spread functions derived from the ERF and DTF methods were in excellent agreement. Uncorrected veiling glare reduced contrast and the image intensity in CBCT reconstruction, near the phantom periphery (by 67 Hounsfield units in a 20 cm-in-diameter water phantom) and (to a smaller degree) near inhomogeneities. Use of the bow-tie filter mitigated these effects. Veiling glare also resulted in about 10%-15% overestimation of the scatter-to-primary ratio when measured with the beam-stop or beam-stop array method.Conclusions: The long tails of the detector PSF were found to have a modest dependence of beam spectrum, which is reflected on the MTF curve LFD. Our findings show that uncorrected veiling glare can affect quantitative accuracy and contrast in CBCT imaging, based on flat panel imager. In addition, it results in overestimation of the scatter-to-primary ratio, measured with the beam-stop methods. (c) 2012 American Association of Physicists in Medicine. [http://dx.doi.org/10.1118/1.4747260]