任意比工况下基于像素的改进投影矩阵计算方法

Improved Projection Matrix Computation Method Based on Pixel under the Working Condition with an Arbitrary Radio

针对CT图像重建代数类算法中关键的投影矩阵,提出1种在像素尺寸与探测器间隔为任意比工况下的计算方法。在各投影方向上,根据该比值确定与像素可能相交的最大射线条数,从每个像素出发直接定位到可能与其相交的射线编号,基于射线与像素相交拓扑模型计算的交线长度作为投影矩阵对应元素。利用MATLAB及SNARK09平台仿真验证本文方法的有效性,结果表明:在任意比工况下计算投影矩阵有其实际意义;本文方法能在任意比工况下有效地计算投影矩阵,其效率相对于传统的Siddon方法提高4倍以上;本文方法能有效地应用于SART图像重建算法中。

For the project matrix which is critical in CT image algebraic reconstruction algorithms, a computation method for dealing with an arbitrary pixel-size to detector-spacing ratio case was presented. In each projection direction, the maximum number of the lines that may intersect a pixel was determined according to the ratio, and then the lines were located based on each pixel, and finally intersection lengths were computed to be assigned to the related project matrix elements according to the intersection topological model. MATLAB and SNARK09 software platforms were employed for simulation experiments to justify effectiveness of the method. Results show it is significant to compute the project matrix in the arbitrary ratio case, and that in the arbitrary ratio working condition a projection matrix can be effectively computed by the method, which is four times efficient than the classical Siddon algorithm, and that the method could be effectively used in SART image reconstruction algorithms.