一种适用于小视野的快速精确重建算法  

An Efficient Theoretically-Exact Algorithm for Small Field of View

作　　者：姜永生[1] 于燕燕[1] 

机构地区：[1]东北大学中荷生物医学与信息工程学院,辽宁沈阳110819

出　　处：《东北大学学报（自然科学版）》2012年第12期1714-1717,共4页Journal of Northeastern University(Natural Science)

基　　金：国家自然科学基金资助项目(51006021)

摘　　要：分析了反投影滤波(BPF)精确重建算法在重建小视野图像重建中存在的速度缓慢问题,提出一种适用于小视野的快速精确重建算法.该算法通过多尺度采样概念重新定义PI平面数据,利用两次不同尺度PI平面数据的插值获得最终结果,避免了原始算法在PI平面的无用计算.实验结果表明,该算法能够在保证重建图像质量的前提下,大大降低PI平面的计算量,从而提高BPF精确重建算法的重建速度,有效地解决了精确重建算法在小视野重建中的无用计算问题,为精确重建算法应用于实际提供了条件.The problems of back projection filtered（BPF） reconstruction in small field of view were analyzed,and an efficient theoretically-exact reconstruction algorithm for small field of view was proposed.With the proposed algorithm,useless computation in reconstructing PI plane was avoided by redefining the sampling method of PI plane with multi-scale method,and interpolating the two different scale PI plane data.The algorithm depressed the computation of PI plane and got the same results as original BPF algorithm,which accelerated the BPF reconstruction for solving the inutility computation of BPF algorithm in small field of view reconstruction,and also provided a good chance for clinical use.

关 键 词：反投影滤波(BPF) 精确重建 多尺度 PI平面 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]