A review on generative based methods for MRI reconstruction

Abstract

Magnetic resonance imaging (MRI) is one of the most important methods for clinical diagnosis. However, the main drawback of MRI is the long imaging time, which will cause the moving artifact by patient movements. With the rapid development of the computing power of computer, deep learning is widely used in computer vision, natural language processing, visual recognition and so on. Meanwhile, a large number of reconstruction methods based on deep learning have also emerged. Recently, many generative models have been proposed to solve the perception quality problem that existed in fast MRI images. In this paper, we manage to survey the motivations and reconstruction strategies of generative-based methods published in journals and conferences over the past five years. First, the background and theoretical basis of MRI reconstruction are introduced. Secondly, the application of generative-based methods in MRI reconstruction field is comprehensively summarized and analyzed, including Generative Adversarial Network (GAN), Variational Autoencoder (VAE) and VAE-GAN. Then the advantages and disadvantages of the existing generative-based MRI reconstruction methods are discussed. Finally, several publicly available MR image datasets and evaluation metrics are presented, which can provide a reference for researchers and practitioners working in related domains. The conclusions and challenges are also given.

HARA-GAN: Hybrid Attention and Relative Average Discriminator Based Generative Adversarial Network for MR Image Reconstruction

“…The results show that HARA-GAN performs better than other methods including DAGAN [5], RefineGAN [6], and RSCA-GAN [11], according to various quantitative metrics. The SSA [23] shows the improved generator performance of the model and the relative average discriminator [14] improved the discriminator performance.…”

“…DAGAN [5] proposed a novel deep de-aliasing generative adversarial network (DAGAN) for fast CS-MRI reconstruction. In DAGAN the details of the brain cannot be distinguished well.…”

“…Generating high-quality and consistent MR images from the under sampled input MR image were an active and challenging research problem. While previous research has shown promising results in high under sampling rates, the effectiveness of those techniques [5] [6] [7] [8] in low under sampling rates remains limited based on the quantitative evaluation metrics. Currently, Deep learning-based approaches outperform traditional methods such as parallel imaging(PI) and compressed sensing (CS) in MRI reconstruction [9].…”