Denoising of 3-D Magnetic Resonance Images Using a Residual Encoder-Decoder Wasserstein Generative Adversarial Network

by   Maosong Ran, et al.

Structure-preserved denoising of 3-D magnetic resonance images (MRI) is a critical step in the medical image analysis. Over the past years, many algorithms have been proposed with impressive performances. Inspired by the idea of deep learning, in this paper, we introduce a MRI denoising method based on the residual encoder-decoder Wasserstein generative adversarial network (RED-WGAN). Specifically, to explore the structure similarity among neighboring slices, 3-D configuration are utilized as the basic processing unit. Residual autoencoder, combined with deconvolution operations are introduced into the generator network. Furthermore, to alleviate the shortcoming of traditional mean-squared error (MSE) loss function for over-smoothing, the perceptual similarity, which is implemented by calculating the distances in the feature space extracted by a pre-trained VGG-19 network, is incorporated with MSE and adversarial losses to form the new loss function. Extensive experiments are studied to access the performance of the proposed method. The experimental results show that the proposed RED-WGAN achieves superior performance relative to several state-of-art methods in both simulated and clinical data. Especially, our method demonstrates powerful ability in both noise suppression and structure preservation.


Low Dose CT Image Denoising Using a Generative Adversarial Network with Wasserstein Distance and Perceptual Loss

In this paper, we introduce a new CT image denoising method based on the...

Enhanced generative adversarial network for 3D brain MRI super-resolution

Single image super-resolution (SISR) reconstruction for magnetic resonan...

Denoising of 3D magnetic resonance images with multi-channel residual learning of convolutional neural network

The denoising of magnetic resonance (MR) images is a task of great impor...

Adversarial and Perceptual Refinement for Compressed Sensing MRI Reconstruction

Deep learning approaches have shown promising performance for compressed...

On denoising autoencoders trained to minimise binary cross-entropy

Denoising autoencoders (DAEs) are powerful deep learning models used for...

Adversarial Distortion Learning for Medical Image Denoising

We present a novel adversarial distortion learning (ADL) for denoising t...

Please sign up or login with your details

Forgot password? Click here to reset