Mutual Information Estimation Using K-Space Data: An Application For Eddy-Current Distortion Correction In Diffusion Tensor Imaging
Guan X, Lai S, Shi J, Lackey J, Techavipoo U
Thomas Jefferson University
Mutual information between diffusion-weighted and T2-weighted images, used for eddy-current distortion correction in diffusion tensor imaging, is generally estimated using linear interpolation and partial volume methods. These methods create artifacts on MI and reduce the accuracy of the distortion-correction parameters. We developed a method based on the shifting theorem to discover pixel values in any specific coordinates without interpolation. Upsampling was applied to reduce the computational time. The joint probability between two images was smoothed using 2D-Gaussian kernel. The results indicate major improvement in terms of principal-components variances and visual qualities of fractional anisotropy images and brain-fiber-orientation color maps.