scholarly journals CS-GAN for High-Quality Diffusion Tensor Imaging

2020 ◽  
Author(s):  
Ying Cao ◽  
Lihui Wang ◽  
Jianping Huang ◽  
Xinyu Cheng ◽  
Jian Zhang ◽  
...  
Keyword(s):  
Diffusion Tensor Imaging ◽  
Image Quality ◽  
Compressed Sensing ◽  
Spatial Information ◽  
Diffusion Tensor ◽  
High Quality ◽  
Image Domain ◽  
Low Snr ◽  
Sampling Ratio ◽  
Space Data

Abstract Background: Compressed sensing magnetic resonance imaging (CS-MRI) is a promising technique for accelerating MRI speed. However, image quality in CS-MRI is still a pertinent problem. In particular, there is little work on reducing aliasing artefacts in compressed sensing diffusion tensor imaging (CS-DTI), which constitute a serious obstacle to obtaining high-quality images. Method: We propose a CS-DTI de-aliasing method based on conditional generative adversarial (cGAN), called CS-GAN, to tackle de-aliasing problems in CS-DTI with highly undersampled k-space data. The method uses a nested-UNet based generator, a ResNet-based discriminator, and a content loss function defined in both image domain and frequency domain. Result and Concludions: Compared to existing state-of-the-art de-aliasing methods based on deep learning, our method achieves superior imaging quality in terms of both diffusion weighted (DW) image quality and DTI diffusion metrics. Moreover, even at extremely low sampling ratio and low SNR, our method can still reconstruct texture details and spatial information.

scholarly journals Diffusion Tensor Imaging of the Kidneys: Influence of b-Value and Number of Encoding Directions on Image Quality and Diffusion Tensor Parameters

2013 ◽  
Vol 3 ◽  
pp. 53 ◽  
Author(s):  
Natalie C. Chuck ◽  
Günther Steidle ◽  
Iris Blume ◽  
Michael A. Fischer ◽  
Daniel Nanz ◽  
...  
Keyword(s):  
Diffusion Tensor Imaging ◽  
Image Quality ◽  
Fractional Anisotropy ◽  
Diffusion Tensor ◽  
Mean Diffusivity ◽  
Renal Medulla ◽  
Whole Body ◽  
Acquisition Time ◽  
Data Sets ◽  
B Values

Objectives: The purpose of this study was to evaluate to which degree investment of acquisition time in more encoding directions leads to better image quality (IQ) and what influence the number of encoding directions and the choice of b-values have on renal diffusion tensor imaging (DTI) parameters. Material and Methods: Eight healthy volunteers (32.3 y ± 5.1 y) consented to an examination in a 1.5T whole-body MR scanner. Coronal DTI data sets of the kidneys were acquired with systematic variation of b-values (50, 150, 300, 500, and 700 s/mm2) and number of diffusion-encoding directions (6, 15, and 32) using a respiratory-triggered echo-planar sequence (TR/TE 1500 ms/67 ms, matrix size 128 × 128). Additionally, two data sets with more than two b-values were acquired (0, 150, and 300 s/mm2 and all six b-values). Parametrical maps were calculated on a pixel-by-pixel basis. Image quality was determined with a reader score. Results: Best IQ was visually assessed for images acquired with 15 and 32 encoding directions, whereas images acquired with six directions had significantly lower IQ ratings. Image quality, fractional anisotropy, and mean diffusivity only varied insignificantly for b-values between 300 and 500 s/mm2. In the renal medulla fractional anisotropy (FA) values between 0.43 and 0.46 and mean diffusivity (MD) values between 1.8-2.1 × 10-3 mm2/s were observed. In the renal cortex, the corresponding ranges were 0.24-0.25 (FA) and 2.2-2.8 × 10-3 mm2/s (MD). Including b-values below 300 s/mm2, notably higher MD values were observed, while FA remained constant. Susceptibility artifacts were more prominent in FA maps than in MD maps. Conclusion: In DTI of the kidneys at 1.5T, the best compromise between acquisition time and resulting image quality seems the application of 15 encoding directions with b-values between 300 and 500 s/mm2. Including lower b-values allows for assessment of fast diffusing spin components.

scholarly journals Cardiac diffusion tensor imaging based on compressed sensing using joint sparsity and low-rank approximation

2016 ◽  
Vol 24 (s2) ◽  
pp. S593-S599 ◽  
Author(s):  
Jianping Huang ◽  
Lihui Wang ◽  
Chunyu Chu ◽  
Yanli Zhang ◽  
Wanyu Liu ◽  
...  
Keyword(s):  
Diffusion Tensor Imaging ◽  
Compressed Sensing ◽  
Diffusion Tensor ◽  
Low Rank ◽  
Low Rank Approximation ◽  
Joint Sparsity ◽  
Rank Approximation

scholarly journals Highly Accelerated Vessel-Selective Arterial Spin Labelling Angiography using Sparsity and Smoothness Constraints

2019 ◽  
Author(s):  
Sophie Schauman ◽  
Mark Chiew ◽  
Thomas W. Okell
Keyword(s):  
Image Quality ◽  
Compressed Sensing ◽  
Arterial Spin Labeling ◽  
Parallel Imaging ◽  
Spin Labeling ◽  
Dominant Factor ◽  
Image Domain ◽  
In Vivo Experiments ◽  
Scan Time

AbstractPurposeTo demonstrate that vessel-selectivity in arterial spin labeling angiography can be achieved without any scan time penalty or noticeable loss of image quality compared to conventional arterial spin labeling angiography.MethodsSimulations on a numerical phantom were used to assess whether the increased sparsity of vessel-encoded angiograms compared to non-vessel-encoded angiograms alone can improve reconstruction results in a compressed sensing framework. Further simulations were performed to study whether the difference in relative sparsity between non-selective and vessel-selective dynamic angiograms were sufficient to achieve similar image quality at matched scan times in the presence of noise. Finally, data were acquired from 5 healthy volunteers to validate the technique in vivo. All data, both simulated and in vivo, were sampled in 2D using a golden angle radial trajectory and reconstructed by enforcing both image domain sparsity and temporal smoothness on the angiograms in a parallel imaging and compressed sensing framework.ResultsRelative sparsity was established as a primary factor governing the reconstruction fidelity. Using the proposed reconstruction scheme, differences between vessel-selective and non-selective angiography were negligible compared to the dominant factor of total scan time in both simulations and in vivo experiments at acceleration factors up to R = 34. The reconstruction quality was not heavily dependent on hand-tuning the parameters of the reconstruction.ConclusionThe increase in relative sparsity of vessel-selective angiograms compared to non-selective angiograms can be leveraged to achieve higher acceleration without loss of image quality, resulting in the acquisition of vessel-selective information at no scan time cost.

scholarly journals Compressed-Sensing Magnetic Resonance Image Reconstruction Using an Iterative Convolutional Neural Network Approach

2020 ◽  
Vol 10 (6) ◽  
pp. 1902
Author(s):  
Fumio Hashimoto ◽  
Kibo Ote ◽  
Takenori Oida ◽  
Atsushi Teramoto ◽  
Yasuomi Ouchi
Keyword(s):  
Magnetic Resonance ◽  
Image Quality ◽  
Compressed Sensing ◽  
Signal To Noise Ratio ◽  
A Priori ◽  
Similarity Index ◽  
Structural Similarity ◽  
Neural Network Approach ◽  
Priori Information ◽  
Space Data

Convolutional neural networks (CNNs) demonstrate excellent performance when employed to reconstruct the images obtained by compressed-sensing magnetic resonance imaging (CS-MRI). Our study aimed to enhance image quality by developing a novel iterative reconstruction approach that utilizes image-based CNNs and k-space correction to preserve original k-space data. In the proposed method, CNNs represent a priori information concerning image spaces. First, the CNNs are trained to map zero-filling images onto corresponding full-sampled images. Then, they recover the zero-filled part of the k-space data. Subsequently, k-space corrections, which involve the replacement of unfilled regions by original k-space data, are implemented to preserve the original k-space data. The above-mentioned processes are used iteratively. The performance of the proposed method was validated using a T2-weighted brain-image dataset, and experiments were conducted with several sampling masks. Finally, the proposed method was compared with other noniterative approaches to demonstrate its effectiveness. The aliasing artifacts in the reconstructed images obtained using the proposed approach were reduced compared to those using other state-of-the-art techniques. In addition, the quantitative results obtained in the form of the peak signal-to-noise ratio and structural similarity index demonstrated the effectiveness of the proposed method. The proposed CS-MRI method enhanced MR image quality with high-throughput examinations.

scholarly journals Prospective acceleration of diffusion tensor imaging with compressed sensing using adaptive dictionaries

2015 ◽  
Vol 76 (1) ◽  
pp. 248-258 ◽  
Author(s):  
Darryl McClymont ◽  
Irvin Teh ◽  
Hannah J. Whittington ◽  
Vicente Grau ◽  
Jürgen E. Schneider
Keyword(s):  
Diffusion Tensor Imaging ◽  
Compressed Sensing ◽  
Diffusion Tensor
Sign in / Sign up

Export Citation Format

Share Document