论文部分内容阅读
为了提高能谱X-CT重建图像的质量,提出了利用能量加权重建图像x_(bins)~W及可分离抛物面替代法进行基于先验图像和约束压缩感知的能谱X-CT图像重建.利用压缩感知理论、先验图像和优化算法来提高CT重建图像的质量.为了评价所提方法的性能,从重建的各能量段图像精度和噪声特性2个方面比较了3种优化算法及3种先验图像.仿真实验结果表明,对于不同的优化算法,能量加权重建图像xW bins作为先验图像总体性能最佳;对于不同的先验图像,可分离抛物面替代法算法性能最佳.与滤波反投影算法相比,在基于先验图像约束和压缩感知的能谱X-CT图像重建算法中,采用SPS算法进行优化,采用能量加权重建图像作为先验图像,重建得到的各能量段的图像噪声分别降低了80.46%,82.51%,88.08%,每个能量段图像的均方根误差分别下降了15.02%,18.15%和34.11%,相关系数分别提高了9.98%,11.38%和15.94%.
In order to improve the quality of the X-CT image, an energy spectrum X-CT image reconstruction based on priori image and constrained compressive sensing is proposed by using the energy-weighted reconstruction images x bins and the separable paraboloid substitution method. Compressed sensing theory, priori image and optimization algorithm to improve the quality of CT reconstruction image.To evaluate the performance of the proposed method, three kinds of optimization algorithms and three kinds of optimization algorithms are compared from the reconstructed image quality and noise characteristics of each energy segment The experimental results show that, for different optimization algorithms, the energy-weighted reconstruction image xW bins has the best overall performance as a priori image, and the separable paraboloid replacement algorithm has the best performance for different prior images.Compared with the filtered backprojection Compared with the algorithm, the SPS algorithm is used to optimize the X-CT image reconstruction algorithm based on the priori image constraint and compression perception. The energy weight is used to reconstruct the image as a priori image, and the reconstructed image noise of each energy segment is respectively The average root mean square error of each energy segment decreased by 15.02%, 18.15% and 34.11% respectively, and the correlation coefficient increased by 9.98%, 11.38% and 15% respectively .94%.