扩散张量图像去噪算法研究进展

杨黎明, 王远军

Research Progress of Denoising Algorithms for Diffusion Tensor Images

YANG Liming, WANG Yuanjun

图8 利用模拟DWI数据对GL-HOSVD与VST-OSVS算法在4%噪声水平和b值为1 000 s/mm2下的去噪图像、拟合的FA图、彩色编码FA图与无噪图像、噪声图像的视觉比较.数字为对应的RMSE，b值单位为s/mm2

Fig. 8 Visual comparison of the denoised images, fitted FA maps, color-coded FA maps, noise-free images, and noise images of the GL-HOSVD and VST-OSVS algorithms using the simulated DWI data with 4% noise level and b-value of 1 000 s/mm2. The numbers reported are RMSE values, and the b-values are in s/mm2