摘要
实际工程中广泛采用的圆轨道锥束CT重建算法(FDK类算法)中,只有在小锥角(<±5°)的条件下才能较好地重建。为了克服这一缺点,增大圆轨道可重建的锥角范围,推导了圆轨道锥束重建中R adon空间数据缺失与锥角的关系,并基于这一关系提出了基于不同半径的两个同心圆轨道的“改进型误差减小”(im proved error reduction-based,IERB)算法,利用估算的重建误差改善重建结果。数值仿真实验结果表明,IERB算法在锥角≤±10°时都能获得相当好的重建结果,在锥角更大时也可适用于一定的区域。该算法不适用于数据缺失过多的区域。
A computed tomgraphy algorithm was developed that can incorporate larges cone angles in circular cone-beam reconstructions to overcome the shortcomings of widely used algorithms （FDK type algorithms） which only work with small cone angles （〈±5°）. The circular cone-beam reconstruction error was related to the scanning locus radius. An improved error reduction-based method was developed based on two scans in circular orbits with different radii with the estimated reconstruction error. Numeric simulations show that the algorithm works well with cone angles less than 10° and can be utilized in some regions when the cone angle is even larger. The algorithm cannot work in regions with too much incomplete Radon data.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2006年第6期809-812,共4页
Journal of Tsinghua University(Science and Technology)
基金
教育部博士点基金项目(20030003074)
国家自然科学基金资助项目(10135040)
