Improved T2 and diffusion maps from wavelet de-noised magnetic resonance imaging data

By appropriately configuring acquisition parameters, magnetic resonance (MR) images can be influenced by physical factors that include spin-spin (T₂) relaxation rates and diffusion coefficients of the imaged tissues. Usually MR images are the result of a dominating influence of one of these factors to produce T₂-weighted, diffusion-weighted, etc. images. But, increasingly, radiologists and researchers are interested in direct measurement of the physical parameters. This involves the use of two or more images in a calculation that will reveal T₂ maps, diffusion maps, etc. The error introduced by the calculation can be larger than the measurement errors in the original base images so these should be relatively noise-free. Here it is shown how a wavelet based de-noising algorithm can be used to preprocess the base images prior to map calculation to produce improved diffusion and T₂ maps from MR image data