基于Bayes估计的稀疏数据CT重建算法研究

基金项目:

国家自然科学基金(40374048)

详细信息

CT Reconstruction Algorithms with Sparse Radiographs Based on Bayes Estimates

摘要: 通过对稀疏放射数据CT重建问题的分析,提出用Bayes方法重建图像。首先以医疗解剖学为基础,将目标物体的先验信息转化为定量的先验概率密度,即结构先验,将测量所得数据作为似然函数,二者结合得到后验概率密度,通过对其抽样,将均值估算得到期望值作为重建结果。仿真实验证明,与传统CT重建方法相比,Bayes重建放射数据少,重建时间短,图像质量好,抗噪声能力强。
- Bayes估计 /
- CT重建 /
- 先验概率密度
Abstract: A Bayes estimation method is presented via analyzing CT reconstruction with sparse radiograph data.We transform prior information about target object based on medical treatment anatomy to prior probability density and measurement information to likelihood function,and combine them to get posterior probability density.Through sampling and mean estimates,the expected value is taken as the result of CT reconstruction.The simulation results show that,compared to conventional CT reconstruction methods,the Bayes estimation method has less radiation data,short reconstruction time,good-quality images and high noise resistance.
- Bayes method /
- CT reconstruction /
- prior probability density