Image water-fat separation method and water-fat separation system based on magnetic resonance imaging

Abstract

The invention provides an image water-fat separation method and a water-fat separation system based on magnetic resonance imaging. The image water-fat separation method is characterized in that an original image in different echo times can be acquired based on the magnetic resonance imaging, and the field pattern value of every pixel of the original image can be estimated, and then the initial field pattern value of every pixel can be acquired to generate the corresponding field pattern; the field pattern can be processed in a blocked manner, and a plurality of sub-blocks can be acquired; based on the distribution curve of the initial field pattern value of every sub-block, the field pattern value final solution of every pixel can be estimated by adopting the local growth method. By adopting the method and the system provided by the invention, the seed points, the number of which is proper, can be found, and the correct selection of the two solutions (fB, true and fB, swap) of every pixel can be completed by adopting the local growth method, and therefore the water-fat separation can be realized.

Description

technical field [0001] The invention relates to the technology of water and fat in magnetic resonance imaging, in particular to a method and system for separating water and fat in magnetic resonance imaging images. Background technique [0002] In conventional clinical magnetic resonance imaging (Magnetic Resonance Imaging, MRI), fat usually presents high signal, which will cover up the signal generated by water excitation, and it is generally believed that the signal generated by water excitation in tissue contains most of the pathological information. Therefore, the high signal of fat will prevent some lesions from being displayed, and even cause misjudgment of lesions by clinicians. Therefore, the existence of fat signal seriously reduces the value of clinicopathological research of MR signal. The commonly used method now is the multi-echo Dixon (or its variant, extended) water-fat separation method, which uses the frequency chemical shift characteristics between water an...

AI Technical Summary

Problems solved by technology

[0004] Based on the aforementioned various algorithms, there are disadvantages. For example, only one pixel is selected as a seed point on the low-resolution field map image. This single seed point corresponds to a cluster of pixels in the high-resolution image. If the original MR image The tissue in the phase separation is spatially separated, so it is impossible to grow into the phase-separated tissue with this cluster of pixels as the seed point, that is, the field map value in these tissues cannot be correctly estimated; and the latter two methods Different strategies are used to complete the separation of water and fat, but these two methods only get the phasor of the field map, but do not get the real field map value f B

Images