CT系统参数标定模板的设计

Design of parameter calibration template for CT system

本文针对二维CT系统的参数标定问题,在医学数据的基础上进行探究,首先采用普通模板对未知CT系统进行标定,确定了该CT系统旋转中心、探测器单元之间的距离以及X射线的方向;然后基于反投影滤波模型,在MATLAB软件上编写算法实现了CT成像,并运用了更加高效方便的卷积函数代替了滤波过程,以及采用汉宁窗算法减少了傅里叶变换的数据损失;最后分析整个标定与重建模型的精度和稳定性,得出误差主要源于数据的离散性的结论,并进一步总结出标定模板的设计要点,提出了一个高效、易用的标定模板。该模板经过验证,其吸收率函数具有易于拟合的特征。新的模版与标定算法有效提升了标定精度,而且对噪声有着更高的容忍性。

Aiming at the problem of parameter calibration of two-dimensional CT system,this paper explores on the basis of medical data. Firstly,the unknown CT system is calibrated by common template,and the distance between rotating centers,detector units and X-ray direction of the CT system are determined. Then,based on back-projection filtering model,an algorithm is compiled on MATLAB software to realize CT imaging,and more efficient is used. Convolution function replaces the filtering process,and Hanning window algorithm is used to reduce the data loss of Fourier transform. Finally,the accuracy and stability of the whole calibration and reconstruction model are analyzed,and the conclusion that the error mainly comes from the discreteness of data is drawn. The key points of the design of calibration template are further summarized,and an efficient and easy-to-use calibration template is proposed. The absorption function of the template has been verified to be easy to fit. The new template and calibration algorithm can effectively improve the calibration accuracy,and have a higher tolerance to noise.