Development of a next generation X-ray CT with energy spectra of incident photons

2004年我们用蒙特卡罗方法编制了X射线光子输运程序，并对我们提出的光子计数方法的有效性进行了评价。从（1）对比度分辨率的提高、（2）X射线强度的降低、（3）射束硬化效应的消除等方面进行试验。仿真结果表明，与传统的利用X射线能量注量的CT方法相比，该方法可以提高低对比度分辨率约3倍。对于射束硬化效应，我们可以通过对测量的能谱进行加权来消除伪影。并将此方法应用于常规现代X射线系统所能得到的能谱，得到了较好的结果。基于这种考虑，我们设计了一种新的碲化镉光子计数探测器，并利用微聚焦X射线管对探测器的性能进行了评价，在2005年对新研制的X射线探测器进行了多次实验，并将实验结果与蒙特卡罗模拟结果进行了比较。对于测试的材料，我们选择了水，铁，钙，酒精和我们测量透射X射线与光子计数法。在实验中，我们设定了几个阈值来区分X射线的能量，并将实验值与理想值进行了比较。实验数据表明，能量加权计数数据可以提高低密度对比度，同时也可以减少射束硬化伪影。但由于探测器灵敏度的不均匀性，重建的CT图像存在环形伪影，通过对测量计数数据进行适当的插值，可以克服这一问题。

In 2004 we have made a photon transporting code with Monte Carlo method for x-ray photon and evaluated the validity of our proposed photon counting method. The test was performed in terms of (1)improvement of the contrast resolution, (2)reduction of x-ray intensity, and (3)elimination of the beam hardening effect. The results of simulations showed that we could improve the low level contrast resolution about three times compared with the conventional CT method which used the energy fluence of x-rays. For the beam hardening effect we could eliminated the artifacts by weighing the measured energy spectra. And we applied this method to the energy spectra which could be obtained with the conventional modern X-ray system and good results. With this consideration we designed a new photon counting detector which was made of CdTe material and evaluated the performance of the detector with a micro focus x-ray tube.In 2005 we performed several experiment with the newly developed x-ray detector and the performance was compared with the results of Monte Carlo simulation data. For the tested materials we chosen water, Fe, Ca, alcohol and we measured transmitted x-rays with the photon counting method. In this experiment we set several thresholds for discriminating x-rays with their energy and the experiment value was compared with the ideal value. The experiment data showed that we could increase the low density contrast with energy weighted count data and also decrease the beam-hardening artifacts with energy weighted count data. Unfortunately the reconstructed CT images had some ring artifacts which were caused by the non-uniformity of sensitivity of the detector, but this problem could overcome with an appropriate interpolation method for the measured count data.