Denoising of 3D magnetic resonance images using a residual encoder–decoder Wasserstein generative adversarial network

•A MRI denoising method based on the WGAN framework is proposed.•The ideas of residual network and autoencoder are imposed to maintain the structural details and edges, which are clinically important.•With a proper training procedure, our method yields competitive results with several state-of-art m...

Full description

Saved in:
Bibliographic Details
Published in:Medical image analysis Vol. 55; pp. 165 - 180
Main Authors: Ran, Maosong, Hu, Jinrong, Chen, Yang, Chen, Hu, Sun, Huaiqiang, Zhou, Jiliu, Zhang, Yi
Format: Journal Article
Language:English
Published: Netherlands Elsevier B.V 01.07.2019
Elsevier BV
Subjects:
Algorithms
Computer simulation
Deep learning
Feature extraction
Generative adversarial networks
Image analysis
Image denoising
Image processing
Machine learning
Magnetic resonance imaging
Magnetic resonance imaging (MRI)
Medical imaging
NMR
Noise reduction
Nuclear magnetic resonance
Perceptual loss
Preservation
Resonance
Similarity
Wasserstein GAN
Deep learning
Magnetic resonance imaging (MRI)
Perceptual loss
Wasserstein GAN
Image denoising
ISSN:1361-8415, 1361-8423, 1361-8423
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A MRI denoising method based on the WGAN framework is proposed.•The ideas of residual network and autoencoder are imposed to maintain the structural details and edges, which are clinically important.•With a proper training procedure, our method yields competitive results with several state-of-art methods.•The generalization and robustness of our proposed model were carefully sensed by training and testing with different data sources, including simulated and real noise.•Our method is highly computationally fast, and well compatible for parallel implementation on graphic processing units (GPUs). Structure-preserved denoising of 3D magnetic resonance imaging (MRI) images is a critical step in medical image analysis. Over the past few years, many algorithms with impressive performances have been proposed. In this paper, inspired by the idea of deep learning, we introduce an MRI denoising method based on the residual encoder–decoder Wasserstein generative adversarial network (RED-WGAN). Specifically, to explore the structure similarity between neighboring slices, a 3D configuration is utilized as the basic processing unit. Residual autoencoders combined with deconvolution operations are introduced into the generator network. Furthermore, to alleviate the oversmoothing shortcoming of the traditional mean squared error (MSE) loss function, the perceptual similarity, which is implemented by calculating the distances in the feature space extracted by a pretrained VGG-19 network, is incorporated with the MSE and adversarial losses to form the new loss function. Extensive experiments are implemented to assess the performance of the proposed method. The experimental results show that the proposed RED-WGAN achieves performance superior to several state-of-the-art methods in both simulated and real clinical data. In particular, our method demonstrates powerful abilities in both noise suppression and structure preservation. [Display omitted]
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:1361-8415
1361-8423
1361-8423
DOI:10.1016/j.media.2019.05.001