Pith. sign in

REVIEW 4 major objections 6 minor 36 references

Efficient Structurally-Strengthened Generative Adversarial Network for MRI Reconstruction

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A compact GAN beats bigger MRI reconstruction networks.

desk verdict Competent incremental architecture paper for MRI reconstruction whose headline gains over DAGAN/RefineGAN are plausible but not yet verifiable due to underreported baseline training protocol and missing runtime measurements. read the letter →

arxiv 1908.03858 v1 pith:NBFO2UOH submitted 2019-08-11 eess.IV cs.CV

Wenzhong Zhou , Huiqian Du , Wenbo Mei , Liping Fang This is my paper · ORCID
classification eess.IVcs.CV
keywords magneticresonanceimagingcompressedsensinggenerativeadversarialnetworkimagereconstructionresiduallearningstructuralsimilarityconvolutionalautoencoderk-spaceunder-sampling
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes ESSGAN, a generative adversarial network for reconstructing magnetic resonance images from under-sampled k-space data. Its central claim is that ESSGAN achieves higher image quality than the leading deep learning baselines DAGAN and RefineGAN at undersampling rates from 10% to 40% and across radial, Cartesian, and spiral masks, while using only 35.71 million parameters compared with more than 146 million for the two baselines. The quality gains are attributed to three components: strengthened connections that reuse feature maps between two convolutional autoencoders, residual-in-residual blocks embedded in the network's shortcut paths, and an enhanced structural loss that combines multi-scale SSIM with a gradient term. If this is correct, the approach would let fast, high-quality MRI reconstruction run on hardware with limited memory and in tens of milliseconds per slice.

What carries the argument

The load-bearing components are the strengthened connections and the residual-in-residual block. Strengthened connections are shortcut paths that feed the decoder blocks of the first SCAE into the encoder blocks of the second SCAE, rather than only passing the final output forward, so intermediate feature maps are reused between the two autoencoders. The residual-in-residual block contains four convolutions with filter sizes $3\times3$, $1\times1$, $1\times1$, $3\times3$ and two nested residual connections, which deepens the network while keeping parameter count low. These blocks are embedded in the encoder, decoder, and shortcut paths of both SCAEs, and the enhanced structural loss combines $1-\text{MS-SSIM}$ with a gradient penalty to retain texture.

What would settle it

Train ESSGAN, DAGAN, and RefineGAN from the same random seeds on the identical train/validation/test split with the same data augmentation, optimizer hyperparameters, and early-stopping rule, and compare average PSNR at 30% radial undersampling on the same 50 test images; if ESSGAN does not exceed both baselines on that setup, the reported advantage is not robust.

Watch

Extended reading notes

Core claim

The paper's central claim is that a generator made of two strengthened convolutional autoencoders (SCAEs) joined by strengthened connections (SCs), with residual-in-residual blocks (RIRBs) placed in the encoder, decoder, and shortcut paths, reconstructs MR images more accurately than the state-of-the-art GAN baselines while using far fewer parameters. The generator's output is $x_2 = x_0 + x_{C(2,1)}^{out} + x_{C(2,2)}^{out}$, where the two SCAEs contribute their refined feature maps and the input zero-filled image acts as a global residual. The total loss is a weighted sum of the adversarial loss, an $\ell^1$ loss, and an enhanced structural loss $L_{ES} = (1 - \text{MS-SSIM}) + L_{grad}$ that encourages texture and edge preservation. On 50 held-out test images, ESSGAN reports better NMSE and PSNR under radial masks at 10%, 20%, 30%, and 40% undersampling, and better PSNR than both baselines under Cartesian and spiral masks at 30%, while a 256×256 image is reconstructed in tens of milliseconds.

Load-bearing premise

The claim that ESSGAN's architecture is superior to the baselines rests on the assumption that DAGAN and RefineGAN were trained with settings as careful as ESSGAN's; if the published reference implementations were not matched on data split, augmentation, optimizer hyperparameters, and early stopping, the reported quality differences would reflect training imbalance rather than architecture.

Editorial extensions

If this is right

  • The strengthened-connection pattern can be applied to any cascade of convolutional autoencoders to improve feature reuse without sharing weights.
  • Embedding an efficient convolutional block such as the RIRB inside residual shortcuts is a general way to deepen a CNN while controlling parameter growth.
  • At 35.71M parameters, ESSGAN is roughly a quarter the size of the compared baselines, so the architecture is a candidate for memory-constrained or real-time reconstruction hardware.
  • The enhanced structural loss combining MS-SSIM and gradient loss can be transferred to other image restoration tasks that need texture preservation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A fair test of the architectural advantage would retrain DAGAN and RefineGAN with the same data split, augmentation, optimizer schedule, and early-stopping rule; the paper does not report whether this was done, so the reported PSNR gaps should be read as conditional on that assumption.
  • The 'tens of milliseconds' figure is for a single 2D slice; applying the method to clinical volumetric scanning would require measuring the full multi-slice reconstruction pipeline, including data transfer.
  • Because the enhanced structural loss combines an established perceptual metric with an edge penalty, it may behave like a learned multi-scale edge-preserving regularizer, which could be studied in isolation to separate its contribution from the GAN loss.
  • If the parameter reduction generalizes, the same SCAE-with-strengthened-connections design could be retargeted to multi-coil or 3D MRI, where memory is a stronger constraint.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript proposes ESSGAN, a GAN-based MRI reconstruction method whose generator combines two strengthened convolutional autoencoders (SCAEs) with strengthened connections (SCs), residual-in-residual blocks (RIRBs), and an enhanced structural loss formed by combining MS-SSIM loss and gradient loss. The experiments use the MICCAI 2013 dataset, radial/Cartesian/spiral masks at undersampling rates of 10%–40%, and compare against DAGAN and RefineGAN using NMSE, PSNR, and SSIM. The paper claims that ESSGAN achieves higher reconstruction quality with far fewer parameters (35.71M vs. 146.73M and 156.24M) and reconstructs a 256×256 image in tens of milliseconds.

Significance. If the comparisons were properly controlled, the reported gains would be practically useful: a roughly four-fold parameter reduction together with consistent PSNR improvements across masks and undersampling rates would make ESSGAN an attractive baseline for future CS-MRI work. The ablation in Table 4 is a genuine control showing that each proposed component contributes to the full model, and the multi-mask/multi-rate evaluation is broader than in many competing papers. The principal limitation is that the cross-method superiority claim rests on an incompletely reported baseline protocol, no statistical significance testing, and an unsupported runtime claim; these issues are fixable in principle and do not invalidate the architectural idea itself.

major comments (4)
  1. [Section 4.2, Tables 1–3 and 5] The fairness of the baseline comparison is not established. The paper states only that DAGAN and RefineGAN were run using "the source code provided by the authors on GitHub" and does not report whether these models were retrained on the same 15,839/5,050/50 data split, the same data augmentation, the same Adam settings and learning-rate schedule, the same early-stopping criterion, or the same input preprocessing as ESSGAN. Because the central claim of higher PSNR with fewer parameters is entirely empirical, the reported gains (e.g., 2.7 dB over RefineGAN in Table 2) could be a training-protocol artifact rather than an architectural effect. The authors must specify the exact training protocol for all methods and, ideally, retrain all baselines under a single shared pipeline.
  2. [Abstract and Section 4.2] The claim that ESSGAN can reconstruct a 256×256 MR image in tens of milliseconds is not supported by any measurement in the paper. No inference-time table, runtime benchmark, or measurement protocol appears anywhere in Section 4, and the hardware description in Section 4.2 only identifies the training machine. The authors should add actual per-image reconstruction times, including confidence intervals, and state whether preprocessing, k-space-to-image conversion, and GPU transfer are included.
  3. [Section 4.3, Tables 1, 2 and 5] No statistical significance testing is reported for any of the quantitative comparisons. The tables report mean ± standard deviation over 50 test images, and the standard deviations are large enough that the superiority claims are not self-evident; for example, in Table 1 at 30% radial, RefineGAN has PSNR 52.27±2.87 while ESSGAN has 55.47±7.04. Paired tests (e.g., Wilcoxon signed-rank or bootstrap confidence intervals on the paired differences) and, ideally, multiple training seeds are needed to support the claim that ESSGAN is better than the baselines rather than merely different on this particular test sample.
  4. [Table 4] The ablation table supports the qualitative claim that each component contributes, but the differences between ESSGAN-B, ESSGAN-C and the full ESSGAN are small relative to the reported standard deviations (e.g., NMSE 0.018±0.006 vs. 0.015±0.006 for ESSGAN-B vs. ESSGAN). Without repeated runs or significance tests, the statement that "removal of any proposed component will affect the performance" is stronger than the data demonstrate. The authors should add variance over multiple training runs or paired significance tests for the ablation comparisons.
minor comments (6)
  1. [Section 4.1.1] The test set of 50 images is described as "randomly selected" without a seed or selection criterion; using a fixed, published test split or reporting the random seed would make the results reproducible.
  2. [Section 3.2.4] The loss weights α=200 and β=100 are introduced without justification or a sensitivity study; a short analysis of how performance varies with these weights would strengthen the claim that the chosen values are not overfit to the test set.
  3. [Table 3] The parameter counts for DAGAN and RefineGAN are reported without explaining how they were obtained (e.g., from the released checkpoints, from a re-implementation, or from the papers); the counting method and whether generator-only or generator-plus-discriminator parameters are included should be stated.
  4. [Section 3.4.3 and Figure 1] The architecture description would benefit from a precise specification of kernel sizes and channel counts for every convolution, deconvolution, and RIRB operation; Figure 1 is too low-resolution to verify the shortcut topology described in Eqs. (17)–(24).
  5. [Abstract and Introduction] The phrase "state-of-the-art" is used although only DAGAN and RefineGAN are compared; the claim should be narrowed to "previously published deep-learning-based CS-MRI methods compared here" or additional recent baselines should be included.
  6. [Section 4.3.1] The text states that Table 2 shows "more than 2.7dB" improvement under Cartesian and spiral masks; checking the numbers, the improvement over RefineGAN is about 3.1 dB for Cartesian and 2.7 dB for spiral, so the wording should be adjusted to reflect the exact values.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: ESSGAN is an empirical architecture-and-ablation paper whose claims rest on training and benchmark comparisons, not on a derivation chain that reduces to its own inputs.

full rationale

This is an experimental deep-learning paper, not a derivation. The proposed method is defined by a network architecture (SCs, RIRBs, SCAEs) and a hand-composed total loss L_total = L_GAN + αL1 + βLES, where LES = LMS-SSIM + Lgrad. These are design choices with explicit formulas, not quantities claimed to be predicted from first principles, so there is no self-definitional step and no fitted parameter renamed as a prediction. The performance claims are supported by direct comparisons in Tables 1-5 and by an ablation study (Table 4) in which removing SCs, RIRBs, or LES degrades NMSE/PSNR/SSIM, which is a standard control against circularity. The only self-citation by the present authors' group is reference [3] (Han, Du, Gao, Mei), cited in the Introduction merely as one of several classic CS-MRI methods; it is not used to justify any ESSGAN component, loss, or result, so it is not load-bearing. The reported superiority over DAGAN and RefineGAN could be questioned on grounds of benchmark fairness, because Section 4.2 says the implementations came from the authors' GitHub code without stating whether they were retrained with the same data split, augmentation, optimizer schedule, and early stopping; however, that is an empirical-validity and reproducibility concern, not circularity. Likewise, the abstract's 'tens of milliseconds' runtime claim lacks a measurement table, but the absence of evidence is not a circular reduction. The derivation chain, such as it is, is self-contained: the forward equations (16)-(24) define the generator's computation, the losses (7)-(15) define the training objective, and the experimental sections measure the resulting performance against external baselines. No equation is equivalent to another by construction in a way that would force the headline result, and no uniqueness theorem or prior result by these authors is invoked to rule out alternatives.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities; its contributions are network components (SCs, RIRB, SCAE) and a loss combination. The central claim relies on standard deep learning assumptions plus several hand-chosen hyperparameters.

free parameters (4)
  • L1 loss weight alpha = 200
    Chosen by hand to balance adversarial loss and structural loss; affects reconstruction sharpness and is not justified by a selection procedure (Section 3.2.4, Eq. 15).
  • Enhanced structural loss weight beta = 100
    Chosen by hand; controls the contribution of MS-SSIM and gradient loss; not tuned or sensitivity-analyzed (Section 3.2.4, Eq. 15).
  • Number of encoder/decoder blocks M = 4
    Architecture depth in each SCAE; set to 4 in Section 4.2; fixed without ablation across M.
  • Number of filters fnum = 64
    Width of convolutional layers; set in Section 4.2; fixed without ablation.
assumptions (5)
  • domain assumption MR images lie on or near a low-dimensional manifold, and a CNN can learn the mapping from the zero-filled image manifold to the fully-sampled image manifold.
    Stated in Section 2.1 as the reason deep learning methods succeed; if false for the clinical distribution, the learned reconstruction does not generalize.
  • standard math The adversarial game in Eq. (1) with the optimal discriminator Eq. (2) leads to minimization of the Jensen-Shannon divergence Eq. (3), so the generator can match the real data distribution.
    This is the standard GAN theory from Goodfellow et al. [8], invoked in Section 2.2 without proof.
  • domain assumption The zero-filled image x0 = F^H y is a sufficient input representation; no additional k-space consistency layer is needed.
    Problem formulation in Section 3.1; the network never enforces data consistency in k-space, unlike DeepCascade or DAGAN variants. If k-space information must be enforced, reconstruction may hallucinate.
  • domain assumption The MICCAI 2013 dataset (100 training datasets, 50 test images) is representative and the train/valid/test split contains no overlap.
    Section 4.1.1; the claim is tested only on this dataset, so generalization to other anatomies, contrasts, or field strengths is not established.
  • ad hoc to paper The chosen loss weights alpha=200 and beta=100 are appropriate and do not overfit the test set.
    No sensitivity analysis is given; results may depend on these hand-chosen values.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Efficient Structurally-Strengthened Generative Adversarial Network for MRI Reconstruction." pith.science (2026). https://pith.science/paper/NBFO2UOH

@misc{pith2026190803858,
  author       = {Pith},
  title        = {Pith review of: Efficient Structurally-Strengthened Generative Adversarial Network for MRI Reconstruction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NBFO2UOH}},
  note         = {Machine review of arXiv:1908.03858}
}
read the original abstract

Compressed sensing based magnetic resonance imaging (CS-MRI) provides an efficient way to reduce scanning time of MRI. Recently deep learning has been introduced into CS-MRI to further improve the image quality and shorten reconstruction time. In this paper, we propose an efficient structurally strengthened Generative Adversarial Network, termed ESSGAN, for reconstructing MR images from highly under-sampled k-space data. ESSGAN consists of a structurally strengthened generator (SG) and a discriminator. In SG, we introduce strengthened connections (SCs) to improve the utilization of the feature maps between the proposed strengthened convolutional autoencoders (SCAEs), where each SCAE is a variant of a typical convolutional autoencoder. In addition, we creatively introduce a residual in residual block (RIRB) to SG. RIRB increases the depth of SG, thus enhances feature expression ability of SG. Moreover, it can give the encoder blocks and the decoder blocks richer texture features. To further reduce artifacts and preserve more image details, we introduce an enhanced structural loss to SG. ESSGAN can provide higher image quality with less model parameters than the state-of-the-art deep learning-based methods at different undersampling rates of different subsampling masks, and reconstruct a 256*256 MR image in tens of milliseconds.

Figures

Figures reproduced from arXiv: 1908.03858 by the authors.

Figure 1
Figure 1. The overall architecture of the proposed ESSGAN: the proposed strengthened gen [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
Figure 2
Figure 2. The different subsampling masks with an undersampling rate of 30%. [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. (a) and (b) denote the test results (NMSE / PSNR) of the convergence and stability [PITH_FULL_IMAGE:figures/full_fig_p019_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Image quality comparison using radial subsampling mask with an undersampling [PITH_FULL_IMAGE:figures/full_fig_p020_4.png]
Figure 5
Figure 5. Figure 5: Image quality comparison using radial subsampling mask with an undersampling [PITH_FULL_IMAGE:figures/full_fig_p020_5.png]
Figure 6
Figure 6. Figure 6: Image quality comparison using cartesian subsampling mask with an undersampling [PITH_FULL_IMAGE:figures/full_fig_p021_6.png]
Figure 7
Figure 7. Figure 7: Image quality comparison at an undersampling rate of 30% of radial subsampling [PITH_FULL_IMAGE:figures/full_fig_p021_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 20 canonical work pages

  1. [1]

    Lustig, D

    M. Lustig, D. L. Donoho, J. M. Santos, J. M. Pauly, Compressed sensing mri, IEEE signal processing magazine 25 (2008) 72–82. doi:10.1109/MS P.2007.914728

    arXiv 2008

  2. [2]

    C. Chen, J. Huang, Compressive sensing mri with wavelet tree sparsity, in: F. Pereira, C. J. C. Burges, L. Bottou, K. Q. Weinberger (Eds.), Ad- vances in Neural Information Processing Systems 25, Curran Associates, Inc., 2012, pp. 1115–1123. URL http://papers.nips.cc/paper/4630-compressive-sensing-mri -with-wavelet-tree-sparsity.pdf

    work page 2012

  3. [3]

    Y. Han, H. Du, X. Gao, W. Mei, Mr image reconstruction using cosup- port constraints and group sparsity regularisation, Iet Image Processing 11 (2017) 155–163. doi:10.1049/iet-ipr.2016.0615

    work page arXiv 2017

  4. [4]

    B. Zhao, J. P. Haldar, A. G. Christodoulou, Z.-P. Liang, Image reconstruc- tion from highly undersampled (k, t)-space data with joint partial separa- bility and sparsity constraints, IEEE Transactions on Medical Imaging 31 (2012) 1809–1820. doi:10.1109/TMI.2012.2203921

    work page arXiv 2012

  5. [5]

    K. H. Jin, D. Lee, J. C. Ye, A general framework for compressed sensing and parallel mri using annihilating filter based low-rank hankel matrix, IEEE Transactions on Computational Imaging 2 (2016) 480–495. doi: 10.1109/TCI.2016.2601296

    arXiv 2016

  6. [6]

    S. G. Lingala, E. D. Y. Hu, M. Jacob, Accelerated dynamic mri exploiting sparsity and low-rank structure: k-t slr, IEEE Transactions on Medical Imaging 30 (2011) 1042–1054. doi:10.1109/TMI.2010.2100850. 24

    arXiv 2011

  7. [7]

    K. He, X. Zhang, S. Ren, J. Sun, Deep residual learning for image recogni- tion, in: The IEEE Conference on Computer Vision and Pattern Recogni- tion (CVPR), 2016

    work page 2016

  8. [8]

    Goodfellow, J

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, Y. Bengio, Generative adversarial nets, in: Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, K. Q. Weinberger (Eds.), Advances in Neural Information Processing Systems 27, Curran As- sociates, Inc., 2014, pp. 2672–2680. URL http://papers.nips.cc/paper/5423-gene...

    work page 2014

Show all 36 references
  1. [9]

    Isola, J.-Y

    P. Isola, J.-Y. Zhu, T. Zhou, A. A. Efros, Image-to-image translation with conditional adversarial networks, in: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017

    2017

  2. [10]

    J.-Y. Zhu, T. Park, P. Isola, A. A. Efros, Unpaired image-to-image trans- lation using cycle-consistent adversarial networks, in: The IEEE Interna- tional Conference on Computer Vision (ICCV), 2017

    2017

  3. [11]

    Ledig, L

    C. Ledig, L. Theis, F. Huszar, J. Caballero, A. Cunningham, A. Acosta, A. Aitken, A. Tejani, J. Totz, Z. Wang, W. Shi, Photo-realistic single im- age super-resolution using a generative adversarial network, in: The IEEE Conference on Computer Vision and Pattern Recognition (CV...

    2017

  4. [12]

    X. Wang, K. Yu, S. Wu, J. Gu, Y. Liu, C. Dong, Y. Qiao, C. Change Loy, Esrgan: Enhanced super-resolution generative adversarial networks, in: The European Conference on Computer Vision (ECCV) Workshops, 2018

    2018

  5. [13]

    S.-J. Park, H. Son, S. Cho, K.-S. Hong, S. Lee, Srfeat: Single image super- resolution with feature discrimination, in: The European Conference on Computer Vision (ECCV), 2018

    2018

  6. [14]

    J. Chen, J. Chen, H. Chao, M. Yang, Image blind denoising with generative 25 adversarial network based noise modeling, in: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018

    2018

  7. [15]

    G. Yang, H. D. S. Yu, G. Slabaugh, P. L. Dragotti, X. Ye, F. Liu, S. Arridge, J. Keegan, Y. Guo, D. Firmin, Dagan: Deep de-aliasing generative adversarial networks for fast compressed sensing mri recon- struction, IEEE Transactions on Medical Imaging 37 (2018) 1310–1321. doi:1...

    2018

  8. [16]

    T. M. Quan, T. Nguyen-Duc, W.-K. Jeong, Compressed sensing mri re- construction using a generative adversarial network with a cyclic loss, IEEE Transactions on Medical Imaging 37 (2018) 1488–1497. doi: 10.1109/TMI.2018.2820120

    2018

  9. [17]

    S. Wang, Z. Su, L. Ying, X. Peng, S. Zhu, F. Liang, D. Feng, D. Liang, Accelerating magnetic resonance imaging via deep learning, in: Inter- national Symposium on Biomedical Imaging, 2016. doi:10.1109/ISBI .2016.7493320

    2016

  10. [18]

    y. yang, J. Sun, H. Li, Z. Xu, Deep admm-net for compressive sensing mri, in: D. D. Lee, M. Sugiyama, U. V. Luxburg, I. Guyon, R. Garnett (Eds.), Advances in Neural Information Processing Systems 29, Curran Associates, Inc., 2016, pp. 10–18. URL http://papers.nips.cc/paper/640...

    2016

  11. [19]

    Schlemper, J

    J. Schlemper, J. Caballero, J. V. Hajnal, A. Price, D. Rueckert, A deep cascade of convolutional neural networks for dynamic mr image recon- struction, IEEE Transactions on Medical Imaging 37 (2018) 491–503. doi:10.1109/TMI.2017.2760978

    2018

  12. [20]

    Ravishankar, Y

    S. Ravishankar, Y. Bresler, Mr image reconstruction from highly undersam- pled k-space data by dictionary learning, IEEE Transactions on Medical Imaging 30 (2011) 1028–1041. doi:10.1109/TMI.2010.2090538. 26

    2011

  13. [21]

    Ronneberger, P

    O. Ronneberger, P. Fischer, T. Brox, U-net: Convolutional net- works for biomedical image segmentation, in: Medical Image Com- puting and Computer-Assisted Intervention, 2015. doi:10.1007/ 978-3-319-24574-4_28

    2015

  14. [22]

    H. Zhao, O. Gallo, I. Frosio, J. Kautz, Loss functions for image restoration with neural networks, IEEE Transactions on Computational Imaging 3 (2017) 47–57. doi:10.1109/TCI.2016.2644865

    2017

  15. [23]

    C.You, Q. Yang, H. Shan, L. Gjesteby, G. Li, S. Ju, Z. Zhang, Z. Zhao, Y. Zhang, W. Cong, G. Wang, Structurally-sensitive multi-scale deep neu- ral network for low-dose ct denoising, IEEE Access 6 (2018) 41839–41855. doi:10.1109/ACCESS.2018.2858196

    2018

  16. [24]

    Z. Wang, A. C. Bovik, H. R. Sheikh, E. P. Simoncelli, Image quality assess- ment: from error visibility to structural similarity, IEEE Transactions on Image Processing 13 (2004) 600–612. doi:10.1109/TIP.2003.819861

    2004

  17. [25]

    Z. Wang, E. P. Simoncelli, A. C. Bovik, Multiscale structural similarity for image quality assessment, in: Asilomar Conference on Signals, Systems and Computers, 2003. doi:10.1109/ACSSC.2003.1292216

    2003 arXiv

  18. [26]

    J. Ma, W. Yu, P. Liang, C. Li, J. Jiang, Fusiongan: A generative adversarial network for infrared and visible image fusion, Information Fusion 48 (2019) 11–26

    2019

  19. [27]

    Zhuang, Laddernet: Multi-path networks based on u-net for medical image segmentation (2018)

    J. Zhuang, Laddernet: Multi-path networks based on u-net for medical image segmentation (2018). arXiv:arXiv:1810.07810

    2018 arXiv

  20. [28]

    Lustig, D

    M. Lustig, D. Donoho, J. M. Pauly, Sparse mri: the application of com- pressed sensing for rapid mr imaging, Magnetic Resonance in Medicine 58 (2007) 1182–1195. doi:10.1002/mrm.21391

    2007 doi

  21. [29]

    R. G. Baraniuk, Compressive sensing, IEEE Signal Processing Magazine 24 (2007) 118–121. doi:10.1109/MSP.2007.4286571. 27

    2007

  22. [30]

    J. Yang, Y. Zhang, W. Yin, A fast alternating direction method for tvl1-l2 signal reconstruction from partial fourier data, IEEE Journal of Selected Topics in Signal Processing 4 (2010) 288–297. doi:10.1109/JSTSP.2010. 2042333

    2010 doi

  23. [31]

    X. Qu, D. Guo, B. Ning, Y. Hou, Y. Lin, S. Cai, Z. Chen, Undersampled mri reconstruction with patch-based directional wavelets, Magnetic Resonance Imaging 30 (2012) 964–977

    2012

  24. [32]

    X. Qu, Y. Hou, F. Lam, D. Guo, J. Zhong, Z. Chen, Magnetic resonance im- age reconstruction from undersampled measurements using a patch-based nonlocal operator, Medical Image Analysis 18 (2014) 843–856

    2014

  25. [33]

    Pawar, G

    K. Pawar, G. Egan, J. Zhang, Multichannel compressive sensing mri using noiselet encoding, PLOS ONE 10 (2015) e0126386. doi:10.1371/journa l.pone.0126386

    2015 doi

  26. [34]

    E. M. Eksioglu, Decoupled algorithm for mri reconstruction using nonlocal block matching model: Bm3d-mri, Journal of Mathematical Imaging and Vision 56 (2016) 430–440. doi:10.1007/s10851-016-0647-7

    2016 doi

  27. [35]

    Mardani, E

    M. Mardani, E. Gong, J. Y. Cheng, S. Vasanawala, G. Zaharchuk, M. Al- ley, N. Thakur, S. Han, W. Dally, J. M. Pauly, L. Xing, Deep genera- tive adversarial networks for compressed sensing automates mri (2017). arXiv:arXiv:1706.00051

    2017 arXiv

  28. [36]

    S. Yu, H. Dong, G. Yang, G. Slabaugh, P. L. Dragotti, X. Ye, F. Liu, S. Ar- ridge, J. Keegan, D. Firmin, Y. Guo, Deep de-aliasing for fast compressive sensing mri (2017). arXiv:arXiv:1705.07137. 28

    2017 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.