各向异性边缘检测导引的CT稀疏角度重建

Anisotropic edge guided CT reconstruction from sparse views

目的:提出一种CT稀疏角度重建方法,用以降低CT的辐射剂量。方法:基于图像各向异性的边缘检测信息,在图像重建过程中更新图像总变差各加权分量的加权权重,然后采用图像总变差最小化方法重建图像。结果:计算机模拟结果表明,与传统的总变差最小化方法相比,该方法可以更好地重建图像边缘,且边缘检测的结果会影响重建效果。结论:采用该方法可以增强图像边缘,提高重建图像的分辨率,改善重建图像的质量。

Objective To develop a reconstruction algorithm for sparse view computed tomography（CT） to decrease the radiation dose. Methods Firstly, the weights of the adding elements of the image total variation were updated according to the anisotropic edge detection information. Then the total variation minimization（TV minimization） algorithm was implemented to reconstruct the image. Results Computer simulation showed that the proposed algorithm could reconstruct the edge area much better than the conventional TV minimization. The precision of edge detection affected the reconstruction results. Conclusion The proposed algorithm can enhance the image edge and improve the resolution and quality of reconstructed images.