Gradient magnetic field shimming method based on regularization magnetic field distribution image reconstruction

Abstract

The invention discloses a gradient magnetic field shimming method based on regularization magnetic field distribution image reconstruction. Compared with an existing phase difference imaging gradient magnetic field shimming technology, the image reconstruction optimization method obtains an initial field pattern through three times of phase difference scanning imaging, a penalized likelihood estimation function of field pattern data is built by a penalized likelihood estimation method with smooth and prior potential functions, function minimization is carried out through optimization transformation, and a final field pattern is obtained. By means of the method, on one hand, continuity of pixels can be effectively considered, and noise and image burrs are restrained; on the other hand, the target shape of the data can be estimated through the prior potential function, it is avoided that the direction is error in the optimization transformation process, and a good correcting effect is obtained for image distortion and winding caused by inhomogeneous magnetic fields. The method can ensure the integrity of the field pattern and improve the signal to noise ratio, so that a certain magnetic field shimming effect is obtained even under extremely poor magnetic field homogeneity conditions like a cold field.

Description

technical field [0001] The invention relates to a gradient shimming technology of a nuclear magnetic resonance spectrometer, and more particularly to a gradient shimming method based on regularized magnetic field distribution image reconstruction on a nuclear magnetic resonance spectrometer. Background technique [0002] For modern nuclear magnetic resonance spectrometers, the uniformity of the static magnetic field directly affects the resolution of the collected signals. In order to improve the uniformity of the static magnetic field, most spectrometers use a series of shim coils with current to generate a specific spatial magnetic field to eliminate the inhomogeneous component of the static magnetic field (active shimming). Due to the complexity of the orthogonal design of the shimming coil and the distribution of the spatial magnetic field, the traditional manual shimming method often requires too much human intervention by the experimenter. The cumbersome operation and ...

AI Technical Summary

Problems solved by technology

There are two problems in this method of directly fitting the field map data through the phase difference: 1) The influence of noise is not considered

Generally speaking, in the fitting calculation of the field diagram, the noise will be approximately ignored, but in the actual measurement process, it is real or even very large (such as the case where the deuterium content of the sample is very weak), which will often make the image like Poor quality, especially for imaging regions with low spin density, tends to make weaker useful signals buried in noise

2) Not suitable for extremely poor magnetic field conditions

This method can effectively correct the discontinuity in the adjacent area of ​​the field image pixel to make it smooth, but when the magnetic field in this area is too uneven, the distorted and entangled signal cannot be restored to an ideal state, and sometimes even the optimization The transformation process makes it more severe along the direction of the "distorted" point

Images