摘要
为了克服磁共振图像重构精度低的问题,方便医生诊断与治疗,提出一种将组稀疏残差去噪和近似消息传递相结合的磁共振图像重构算法。在基于迭代软阈值的去噪近似消息传递(D-AMP)重构算法中,滤波的去噪算法将使用基于组稀疏残差约束(GSRC)的图像去噪实现。实验结果表明,基于组稀疏残差去噪的磁共振图像重建算法可有效缓解重建图像局部细节信息损失量大的问题,提高了图像重建性能,具有良好的鲁棒性。
In order to overcome the problem of low accuracy of magnetic resonance image reconstruction and facilitate the diagnosis and treatment of doctors,a magnetic resonance image reconstruction algorithm combing group sparse residual denoising and approxi⁃mate message passing is proposed.In the reconstruction algorithm of denoising-based approximate message passing(D-AMP)based on iterative soft threshold,the filtering denoising algorithm uses group sparse residual constraint(GSRC).The experimental results show that the MR image reconstruction algorithm based on group sparse residual denoising can effectively alleviate the problem of large loss of local detail information in the reconstructed image,improves the image reconstruction performance and has good robustness.
作者
袁小君
李杨
杨晓城
蒋明峰
YUAN Xiao-jun;LI Yang;YANG Xiao-cheng;JIANG Ming-feng(School of Information Science and Technology,Zhejiang Sci-Tech University,Hangzhou 310018,China)
出处
《软件导刊》
2021年第1期209-213,共5页
Software Guide
基金
浙江省数理医学学会联合基金重点项目(LSZ19F010001)。
关键词
磁共振
图像去噪
组稀疏残差
去噪近似消息传递
magnetic resonance imaging
image denoising
group sparse residual
denoising-based approximate message passing
