Fast magnetic resonance imaging method based on residual U-net convolutional neural network

Abstract

The invention discloses a fast magnetic resonance imaging method based on residual U-net convolutional neural network and the method comprises three steps of preparing training data, carrying out training based on the residual U-net convolutional neural network, and carrying out image reconstruction based on the residual U-net convolutional neural network. By adding the residual module into the U-net convolutional neural network, the problems of gradient disappearance, overfitting, low convergence speed and the like of the U-net convolutional neural network can be solved, and the quality of rapid MRI imaging based on the U-net convolutional neural network is improved.

Description

technical field [0001] The invention belongs to the field of magnetic resonance imaging, and relates to a fast magnetic resonance imaging method based on a residual U-net convolutional neural network. Background technique [0002] In the past 20 years, Magnetic Resonance Imaging (MRI) has developed rapidly due to its high soft tissue resolution and no ionizing radiation damage to the human body. However, due to the slow imaging speed of MRI, the physiological movement of the subject during the imaging process often causes imaging artifacts, which makes it difficult to meet the requirements of real-time imaging. Therefore, how to speed up the imaging speed of MRI is one of the hot spots in the research of MRI theory and technology. one. [0003] Researchers often shorten the data acquisition time of MRI from three aspects. One is to improve the performance of MRI hardware and enhance the main magnetic field strength and gradient switching speed of the MRI scanner. The magne...

AI Technical Summary

Problems solved by technology

[0010] The present invention aims at the deficiencies of the existing U-net convolutional neural network in the magnetic resonance fast imaging method, improves the existing U-net convolutional neural network, and adds 4 The residual module solves the problem of gradient disappearance during backpropagation of the U-net convolutional neural network, and improves the image reconstruction quality of MRI undersampled data; in addition, there is a problem of excessive swing in the update of the optimized loss function, The present invention adopts Adam (AdaptiveMoment Estimation) optimization algorithm to replace conventional SGD (Stochastic Gradient Descent) optimization algorithm, can further accelerate the convergence speed of U-net convolutional neural network, and can effectively prevent the problem of premature end of training; network parameter initialization Use the initialization method of migration learning to reduce overfitting; use the polynomial decay strategy to make the learning rate decrease smoothly, and decrease faster as the number of rounds increases

Images