A MRI Diffusion Weighted Imaging Method Based on Deep Learning and Data Self-Consistency

Abstract

The invention discloses a magnetic resonance diffusion weighted imaging method based on deep learning and data self-consistency, which relates to the field of magnetic resonance diffusion weighted imaging; it includes: 1: through a network composed of a data self-consistent layer, a CNN network and a phase constraint layer After the modules are repeatedly superimposed to build the network, execute the sequence including pre-scan and navigation echo to determine the input data and training mark data of the network; 2: take the training mark data as the target, and input the image corresponding to the input data into the constructed network through reverse Propagate training to obtain the input-output mapping relationship of the network; 3: Execute a sequence including pre-scanning and no navigation echo to obtain imaging signals and data self-consistent convolution kernels, and input them into the network to obtain output images through forward propagation mapping to complete reconstruction; solve It solves the problem of low imaging resolution due to the limitations of various factors based on traditional diffusion weighted imaging, and achieves the effect of improving network learning ability and imaging resolution.

Description

technical field [0001] The invention relates to the field of magnetic resonance diffusion weighted imaging, in particular to a magnetic resonance diffusion weighted imaging method based on deep learning and data self-consistency. Background technique [0002] Magnetic resonance diffusion imaging technology is a new technology that relies on the random movement of water molecules in the body to provide image contrast. The diffusion of water molecules in tissues conforms to a random thermal motion model, and the magnitude and direction of diffusion are affected by biofilms and biomacromolecules in tissues. When the gradient magnetic field exists, the diffusion motion of water molecules will cause the dephasing of the magnetization vector, resulting in the decrease of the magnetic resonance signal. The degree of MR signal reduction depends on the tissue type, structure, physical and physiological state and microenvironment. In the above process, the gradient magnetic field sp...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is: the present invention discloses a magnetic resonance diffusion weighted imaging method based on deep learning and data self-consistency, which solves the problem of low imaging resolution caused by the limited parallel imaging performance factors of traditional diffusion weighted imaging

Images