Reduction of Gibbs artifacts in magnetic resonance imaging based on Convolutional Neural Network

Wang, Yida; Song, Yang; Xie, Honglan; Li, Wenjing; Hu, Bingwen; Yang, Guang

doi:10.1109/cisp-bmei.2017.8302197

Cited by 17 publications

(21 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep learning for artifact removal in MRI has also been demonstrated for the removal of ghosting artifacts in MR spectroscopy . It has also been demonstrated for removal of Gibbs ringing as well as the removal of streaking artifacts in undersampled radial MRI acquisitions . All of these successful methods are examples of image‐to‐image translation, which refers to the approach of training a deep neural network to predict an image in 1 domain from an image in another domain.…”

Section: Introductionmentioning

confidence: 99%

Conditional generative adversarial network for 3D rigid‐body motion correction in MRI

Johnson

Drangova

2019

Magnetic Resonance in Med

View full text Add to dashboard Cite

westgrid.ca) and Compute Canada Calcul Canada (www.computecanada.ca).Purpose: Subject motion in MRI remains an unsolved problem; motion during image acquisition may cause blurring and artifacts that severely degrade image quality. In this work, we approach motion correction as an image-to-image translation problem, which refers to the approach of training a deep neural network to predict an image in 1 domain from an image in another domain. Specifically, the purpose of this work was to develop and train a conditional generative adversarial network to predict artifact-free brain images from motion-corrupted data.Methods: An open source MRI data set comprising T 2 *-weighted, FLASH magnitude, and phase brain images for 53 patients was used to generate complex image data for motion simulation. To simulate rigid motion, rotations and translations were applied to the image data based on randomly generated motion profiles. A conditional generative adversarial network, comprising a generator and discriminator networks, was trained using the motion-corrupted and corresponding ground truth (original) images as training pairs. Results: The images predicted by the conditional generative adversarial network have improved image quality compared to the motion-corrupted images. The mean absolute error between the motion-corrupted and ground-truth images of the test set was 16.4% of the image mean value, whereas the mean absolute error between the conditional generative adversarial network-predicted and ground-truth images was 10.8% The network output also demonstrated improved peak SNR and structural similarity index for all test-set images. Conclusion: The images predicted by the conditional generative adversarial network have quantitatively and qualitatively improved image quality compared to the motioncorrupted images. K E Y W O R D S conditional generative adversarial networks, convolutional neural networks, deep learning, motion correction, MRI

show abstract

Section: Introductionmentioning

confidence: 99%

Conditional generative adversarial network for 3D rigid‐body motion correction in MRI

Johnson

Drangova

2019

Magnetic Resonance in Med

View full text Add to dashboard Cite

show abstract

“…To further examine the bias and smoothness of neural network images, a uniform phantom and a phantom with varying sized spheres both containing FDG 18 were analyzed. The uniform phantom is shown in Fig.…”

Section: A Whole-body and Whole-body Low-dosementioning

confidence: 99%

“…Neural networks have been applied to improve the image quality in low dose [11]- [14] and limited angle [15], [16] X-ray computed tomography (CT) applications. They have also been applied in magnetic resonance imaging (MRI) to remove Rician noise [17] and to reduce Gibbs artifacts [18]. In the PET low dose imaging domain, they have been used to synthesize normal dose equivalents [19]- [23] utilizing U-Net [24] or ResNet [25] style networks.…”

Section: Introductionmentioning

confidence: 99%

FastPET: Near Real-Time Reconstruction of PET Histo-Image Data Using a Neural Network

Whiteley

Panin

Zhou

et al. 2021

IEEE Trans. Radiat. Plasma Med. Sci.

View full text Add to dashboard Cite

Direct reconstruction of positron emission tomography (PET) data using deep neural networks is a growing field of research. Initial results are promising, but often the networks are complex, memory utilization inefficient, produce relatively small 2-D image slices (e.g., 128 × 128), and low count rate reconstructions are of varying quality. This article proposes FastPET, a novel direct reconstruction convolutional neural network that is architecturally simple, memory space efficient, works for nontrivial 3-D image volumes and is capable of processing a wide spectrum of PET data including low-dose and multitracer applications. FastPET uniquely operates on a histoimage (i.e., image-space) representation of the raw data enabling it to reconstruct 3-D image volumes 67× faster than ordered subsets expectation maximization (OSEM). We detail the FastPET method trained on whole-body and low-dose whole-body data sets and explore qualitative and quantitative aspects of reconstructed images from clinical and phantom studies. Additionally, we explore the application of FastPET on a neurology data set containing multiple different tracers. The results show that not only are the reconstructions very fast, but the images are high quality and have lower noise than iterative reconstructions.

show abstract

“…Gibbs ringing artifact which is caused by partial Fourier (PF) acquisition and zero filling interpolation in MRI data is thoroughly studied by Lee et al [32] and a pipeline was developed for Removal of PF-induced Gibbs ringing (RPG) to remove ringing patterns of different periods by applying the conventional method twice. Deep learning based models [33,34] are also employed for Gibbs ringing artifact removal. Maksim et al [33] proposed an extension of GAS-CNN (Gibbs-ringing Artifact Suppression Convolutional Neural Network) and called it attention-based convolutional neural network for Gibbsringing artifact removal.…”

Section: Related Workmentioning

confidence: 99%

Brain MR Image Enhancement for Tumor Segmentation Using 3D U-Net

Ullah

Ansari

Hanif

et al. 2021

Sensors

View full text Add to dashboard Cite

MRI images are visually inspected by domain experts for the analysis and quantification of the tumorous tissues. Due to the large volumetric data, manual reporting on the images is subjective, cumbersome, and error prone. To address these problems, automatic image analysis tools are employed for tumor segmentation and other subsequent statistical analysis. However, prior to the tumor analysis and quantification, an important challenge lies in the pre-processing. In the present study, permutations of different pre-processing methods are comprehensively investigated. In particular, the study focused on Gibbs ringing artifact removal, bias field correction, intensity normalization, and adaptive histogram equalization (AHE). The pre-processed MRI data is then passed onto 3D U-Net for automatic segmentation of brain tumors. The segmentation results demonstrated the best performance with the combination of two techniques, i.e., Gibbs ringing artifact removal and bias-field correction. The proposed technique achieved mean dice score metrics of 0.91, 0.86, and 0.70 for the whole tumor, tumor core, and enhancing tumor, respectively. The testing mean dice scores achieved by the system are 0.90, 0.83, and 0.71 for the whole tumor, core tumor, and enhancing tumor, respectively. The novelty of this work concerns a robust pre-processing sequence for improving the segmentation accuracy of MR images. The proposed method overcame the testing dice scores of the state-of-the-art methods. The results are benchmarked with the existing techniques used in the Brain Tumor Segmentation Challenge (BraTS) 2018 challenge.

show abstract

Reduction of Gibbs artifacts in magnetic resonance imaging based on Convolutional Neural Network

Cited by 17 publications

References 15 publications

Conditional generative adversarial network for 3D rigid‐body motion correction in MRI

Conditional generative adversarial network for 3D rigid‐body motion correction in MRI

FastPET: Near Real-Time Reconstruction of PET Histo-Image Data Using a Neural Network

Brain MR Image Enhancement for Tumor Segmentation Using 3D U-Net

Contact Info

Product

Resources

About