计算机应用研究2017，Vol.34Issue(7)：2236-2240,5.DOI:10.3969/j.issn.1001-3695.2017.07.069
基于张量奇异值分解的动态核磁共振图像重建
Dynamic MRI reconstruction based on tensor-SVD
徐文 1杨晓梅1
作者信息
- 1. 四川大学 电气信息学院,成都 610065
- 折叠
摘要
Abstract
This paper proposed a new algorithm(t-SVD-TV) to improve the quality of the dynamic magnetic resonance reconstructed image based on tensor singular value decomposition(t-SVD) and total variation sparse model(TV).Algorithm for dynamic magnetic resonance images were sparse and low-rank constraint specification,respectively.The algorithm used the singular value decomposition of the tensor threshold method and total variation method for solving low rank and sparse problem.The experimental results show that the reconstruction and visual effect at the same sampling rate,t-SVD-TV algorithm has better reconstruction quality compared to TV method and k-t SLR,tensor singular value decomposition method.State-of-art image resolution enhancement techniques have improved peak signal-to-noise ratio(PSNR),mean squared error(MSE) and structural similarity(SSIM) with specific application to restruction,denoising and blurring.关键词
MRI/图像重建/张量奇异值分解/动态/全变分Key words
MRI/image reconstruction/tensor singular value decomposition/dynamic/total variation分类
信息技术与安全科学引用本文复制引用
徐文,杨晓梅..基于张量奇异值分解的动态核磁共振图像重建[J].计算机应用研究,2017,34(7):2236-2240,5.
