A Water-fat Separation and Reconstruction Method Based on Single-Scan Spatiotemporal Encoded Magnetic Resonance Imaging

Abstract

A water-fat separation and reconstruction method based on single-scan spatio-temporal encoding magnetic resonance imaging, involving a magnetic resonance imaging method. The method uses a single-scan spatio-temporal encoding imaging sequence to encode and collect water-fat signals, in which a 90-degree linear sweep pulse is combined The spatiotemporal encoding gradient gives the water-lipid signal a quadratic phase during the excitation phase. Based on this characteristic, by selecting appropriate filters to preprocess the collected data, the prior knowledge images of water and lipids can be obtained to guide the separation and reconstruction of water and lipids. In the water-fat separation stage, the prior knowledge image is used as the initial value and weighted information. By solving the super-resolution improvement and edge artifact removal algorithm equations in the water-fat environment, a water map with high resolution and high signal-to-noise ratio can be obtained. and fat chart. The proposed method can effectively achieve water-lipid separation in single-scan spatio-temporally encoded magnetic resonance imaging, while improving the image quality of the separated images and promoting clinical application.

Description

technical field [0001] The present invention relates to a magnetic resonance imaging (Magnetic Resonance Imaging, MRI) method, in particular to a water-fat separation reconstruction method based on single-scan spatio-temporal coding magnetic resonance imaging, which can effectively improve the image resolution of ultra-fast water-fat separation imaging and signal-to-noise ratio. Background technique [0002] The signals detected by magnetic resonance imaging techniques usually come from the signals generated by hydrogen protons in some molecules, such as water and lipids. For the most part, the signal from hydrogen protons in water is of interest for many practical applications. But because fat has a rather short T1 relaxation time, it will appear hyperintense in many important imaging sequences, such as fast spin echo sequences [1] , Spoiled gradient echo sequence [2,3] and the steady-state free precession sequence [4] . Such hyperintensities are very unfavorable for c...

AI Technical Summary

Problems solved by technology

[0003] At present, there are several relatively mature water-fat separation technologies, but they all have the following disadvantages: (1) Chemical shift selective (chemical shift selective, CHESS) fat-compression imaging technology [5,6] Easy to implement and compatible with most sequences, but for B 0 and B 1 Sensitive to field inhomogeneities, which can lead to inaccuracies in liposuction

(2) Spatial-spectral excitation technology [7,8] Directly aim at the water for excitation, which can improve the B 1 resistance to inhomogeneous fields, but is also susceptible to B 0 Effect of inhomogeneous field

The disadvantage is that this technique cannot be applied to T2-weighted imaging or proton density-weighted imaging, and there is some damage to the signal-to-noise ratio

However, since the number of unknowns of the equation to be solved is twice the number of equations, the resolution of the obtained water and fat images will be lost by half, so the super-resolution reconstruction algorithm is crucial for improving the image quality of water-fat separation. important [18-23]

Images