Under-sampling magnetic resonance diffusion spectrum reconstruction method in combination with sparse and low-rank characteristics

Abstract

The invention discloses an under-sampling magnetic resonance diffusion spectrum reconstruction method in combination with sparse and low-rank characteristics, and relates to an under-sampling reconstruction method of magnetic resonance diffusion spectrum. The method comprises generating a Laplace-Fourier joint transformation matrix; establishing an under-sampling reconstruction model in combination with sparse and low-rank characteristics; solving an algorithm based on the under-sampling reconstruction model in combination with sparse and low-rank characteristics; obtaining a recovered diffusion spectrum vector s by the step 3, and obtaining [Rho]s by an operator [Rho] as the ultimately recovered diffusion spectrum. The method comprises generating a Laplace-Fourier joint transform matrix according to experimental parameters; then establishing an under-sampling reconstruction model in combination with sparse and low-rank characteristics; reconstructing the diffusion spectrum vector by an iterative algorithm; and finally transforming the diffusion spectrum vector into the diffusion spectrum. The method reconstructs a complete magnetic resonance diffusion spectrum with a small amountof data, has high reconstruction precision and a good anti-noise ability.

Description

technical field [0001] The invention relates to an under-sampled reconstruction method of a magnetic resonance diffusion spectrum, in particular to a reconstruction method of an under-sampled magnetic resonance diffusion spectrum combined with sparse and low-rank characteristics. Background technique [0002] Magnetic resonance spectroscopy is a very important technology in the field of chemical analysis. Through the collected spectrum, the structure of molecules can be determined, and the metabolites of human tissues can be analyzed. If the sample contains a variety of compounds, coupled with the J-coupling effect of the compound itself, it will often lead to crowding and overlapping of peaks in the one-dimensional magnetic resonance spectrum, which will increase the difficulty of peak assignment and have a negative impact on the analysis and identification of the chemical composition of the sample. cause adverse effects. An effective method is to introduce additional spec...

AI Technical Summary

Problems solved by technology

However, this method only utilizes the sparsity of the diffusion spectrum, fails to utilize the physical model and characteristics of the diffusion spectrum, is susceptible to noise, and generates isolated noise points in the reconstruction, resulting in errors in the reconstruction spectrum

Images