基于线框模型的锥束CT几何参数校正方法  

Geometric parameters calibration method for cone beam CT system based on square-line phantom

作　　者：王珏[1,2] 葛敏雪[1] 蔡玉芳 向前[1] Wang Jue;Ge Minxue;Cai Yufang;Xiang Qian(College of Automation, Chongqing University, Chongqing 400044, China;ICT Research Center, Key Laboratory of Optoelectronic Technology and System of the Ministry of Education, Chongqing University, Chongqing 400044, China)

机构地区：[1]重庆大学自动化学院,重庆400044 [2]重庆大学光电技术及系统教育部重点实验室ICT研究中心,重庆400044

出　　处：《仪器仪表学报》2018年第2期177-184,共8页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(61471070);国家重大科学仪器设备开发专项(2013YQ030629)资助

摘　　要：锥束CT几何参数偏差会导致重建图像畸变或图像模糊。针对工程实际中锥束CT系统初装和应用中的几何参数偏差与漂移,提出了一种改进的基于线框模型的几何参数测量与校正方法。通过对探测器两次不同径向位置的线框模型投影图像进行分析和处理,计算出源-探测器距离及源-旋转中心距离,并确定探测器纵向位置;通过对模型在两次互为180°的投影图像分析,确定探测器横向位置;在此基础上,根据模型参数及其投影图像的测量值求解其他几何参数;最后,用物体投影数据进行带参数重建,获得校正后的重建图像。实验结果表明,该方法能精确求解各几何参数偏差,通过校正可有效减少图像伪影,其主要参数探测器绕其法线旋转角的测量精度可达到0.03°,探测器沿其行和列方向的平移测量精度分别可达到0.03和0.06 pixel,其抗噪性好精度高。同时,该方法降低了对模型安装精度的严格要求,有较高工程实用价值。The reconstructed images often suffer from artifacts caused by misaligned scanner geometry of cone beam CT system（ CBCT）.An improved method based on the square-line phantom is proposed to reduce the geometric artifacts in cone-beam CT images. Firstly,the source-to-detector distance and the source-to-object distance are calculated by analyzing and processing the square-line phantom projected images of the two different radial positions of the detector. The vertical position of the detector is determined. Then,the square-line phantom projection is acquired at the angles with 180 degrees apart,the horizontal position of the detector can be determined. On this basis,the remaining CBCT geometric parameters are resolved according to the model parameters and the calculated values of the projected images. Finally images are reconstruct using modified FDK algorithm. The experimental results demonstrate that the positing accuracy of the rotation angle of the detector is about 0. 03°,and the precisions of orthogonal projection coordinates are 0. 03 and 0. 06 pixel. The proposed method can solve various geometric parameters accurately and reduce image artifacts effectively. Results validate phantom noise immunity,installation simplicity and practicality.

关 键 词：锥束CT 几何参数 线框模型 FDK算法 几何校正 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] TH74[自动化与计算机技术—计算机科学与技术]