Accelerating CS-MRI Reconstruction With Fine-Tuning Wasserstein Generative Adversarial Network

Abstract

Compressed sensing magnetic resonance imaging (CS-MRI) is a time-efficient method to acquire MR images by taking advantage of the highly under-sampled k-space data to accelerate the time consuming acquisition process. In this paper, we proposed a de-aliasing fine-tuning Wasserstein generative adversarial network (DA-FWGAN) for imaging reconstruction of highly under-sampled k-space data in CS-MRI. In the architecture, we used the fine-tuning method for accurate training of the neural network parameters and the Wasserstein distance as the discrepancy measure between the real and reconstructed images. Furthermore, for better preservation of the fine structures in the reconstructed images, we incorporated perceptual loss, image and frequency loss into the loss function for training the network. With experimental results from 3 different sampling schemes and 3 levels of sampling rates, we compared the reconstruction performance of the DA-FWGAN method with other state-of-the-art deep learning methods for CS-MRI reconstruction, including ADMM-Net, Pixel-GAN, and DAGAN. The proposed DA-FWGAN method outperforms all other methods and can provide superior reconstruction with improved peak signalto-noise ratio (PSNR) and structural similarity index measure.INDEX TERMS Fine-tuning, image reconstruction, magnetic resonance image (MRI), Wasserstein generative adversarial network (WGAN).

PIC-GAN: A Parallel Imaging Coupled Generative Adversarial Network for Accelerated Multi-Channel MRI Reconstruction

“…Currently, several novel GAN-based approaches have been proposed for MRI reconstruction. For example, the DA-FWGAN [ 24 ] architecture used a fine-tuning method for training the neural network and the Wasserstein distance as the discrepancy measure between the reference and reconstructed images. SARA-GAN [ 26 ] integrated the self-attention mechanism with relative average discriminator to reconstruct images with more realistic details and better integrity.…”

“…This framework was composed of two consecutive networks, one was used to reconstruct the under-sampled k -space data and the other was used to refine the result. Jiang et al [ 24 ] proposed a de-aliasing fine-tuning Wasserstein generative adversarial network (DA-FWGAN) to perform CS-MRI reconstruction. This approach combines fine-tuning and Wasserstein distance for training.…”

A Review of Deep Learning Methods for Compressed Sensing Image Reconstruction and Its Medical Applications

“…The authors of [129] utilized a U-Net to reconstruct parallel imaging. WGAN [130] was exploited in [131] and three sequentially connected U-Nets were used as the generator.…”

SARA-GAN: Self-Attention and Relative Average Discriminator Based Generative Adversarial Networks for Fast Compressed Sensing MRI Reconstruction

“…At present, GAN and its variants have achieved excellent performance in image-to-image translation (Zhu et al, 2017), image super-resolution (Ledig et al, 2017), and others. In recent years, since its good data representation capabilities, GAN have also been used for MRI fast imaging (Arjovsky et al, 2017;Yang et al, 2017;Jiang et al, 2019;Kwon et al, 2019) and super-resolution (Chen et al, 2018;Lyu et al, 2019;Mahapatra et al, 2019). Yang et al (2017) applied conditional GAN to MRI reconstruction and proposed the De-Aliasing Generative Adversarial Networks (DAGAN) model.…”

“…Wasserstein GAN (Arjovsky et al, 2017) is a variant of the original GAN, by replacing the Jensen-Shannon divergence in the original GAN with Wasserstein distance, it stabilizes the learning process and solves the problem of mode collapse. Jiang et al (2019) proposed a de-aliasing fine-tuning Wasserstein generative adversarial network (DA-FWGAN) for MR imaging reconstruction. The DA-FWGAN could provide reconstruction with improved peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM).…”