中文摘要：

双能X射线计算机断层成像技术（DECT）能够准确地重建物质的有效原子序数和电子密度信息,是一种有效的物质辨别技术。基材料分解法是求解DECT重建问题的主要方法之一。但目前应用此方法得到的重建结果误差较大,并且对于金属伪影抑制能力也较差。本文分析了DECT基材料分解法重建结果的误差来源和形成机制,阐释了其对金属伪影抑制作用差的原因,并推导出了该方法的理论误差计算公式,为分析方便,本文使用了两个单能进行研究。最后,对基材料分解法重建的适用条件提出了建议,一般情况下,该方法对低Z材料误差较小,对高Z材料误差较大。

英文摘要：

Dual-energy computed tomography（DECT）can be used for computing atomic number and electron density.As a result,we can discriminate unknown materials.One approach to deal with DECT is by basis material decomposition method.However,this method leads to a large error and bad suppression of metal artifact.This paper focuses on methodic error analysis of basis material decomposition method.Reasons for bad suppression of metal artifact are explained.Also,this paper derived the theoretical formula of methodic error.We give the applicable conditions of this method.In general,this method leads to a small error for low-Z materials and a large error for high-Z materials.