CT系统模型新模板设计的发现

A Discovery of New Template Design of CT System Model

摘要: 计算机断层扫描(Computed Tomography, CT)技术被广泛地应用于医学、工程领域。模板的形状对CT系统参数标定至关重要，本文针对普通的单椭圆和小圆模板进行了改进，设计了三种新的模板。首先利用Radon变换和R-L滤波获取其接收信息，然后基于卷积反投影重建算法得到其重建图，并利用敏感度、特异度和约登指数对重建效果进行了定量评估。与原模板相比，新模板重建效果更好。此外，为了考察其抗噪性，对原模板和新模板分别加入不同强度的高斯噪声，试验结果表明新模板比原模板具有更好的抗噪性。一个有趣的发现是以多边形轮廓的单参考系几何图形比边缘为弧线的多参考系几何图形为模板对CT重建效果更好。

Abstract: Computed Tomography (CT) technology has been widely used in the field of medicine and engineering. The shape of the template is crucial to the calibration of CT system parameters. In this pa-per, the ordinary single ellipse and small circle template are improved. We design three new templates. The receiving information is obtained by Radon transform and R-L filter firstly, and then their reconstruction images are acquired based on convolution back projection reconstruction algorithm. Furthermore, we make a quantitative assessment of the reconstruction effect by the sensitivity, specificity and Youden index. In addition, in order to investigate the anti-noise, the original template and the new templates are added to the Gauss noise of different strength, respectively. The experimental result shows that the new templates have better noise resistance than the original template. An interesting finding is that the template generated by geometric figure of a single reference frame with polygon outline is better than the template produced by multi reference frame geometry with the arc line for the reconstruction of CT.