A Learnt Half-Quadratic Splitting-Based Algorithm for Fast and High-Quality Industrial Cone-beam CT Reconstruction

Amirkoushyar Ziabari; Aniket Pramanik; Obaidullah Rahman; Singanallur V. Venkatakrishnan

REVIEW 3 major objections 6 minor 16 references

A Learnt Half-Quadratic Splitting-Based Algorithm for Fast and High-Quality Industrial Cone-beam CT Reconstruction

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A learnt half-quadratic splitting algorithm that alternates a CNN denoiser with a conjugate-gradient data-consistency step delivers higher-quality sparse-view cone-beam CT reconstructions than MBIR, U-Net, and a shared-weight variant, in…

desk verdict A memory-efficient PnP/HQS method with unshared per-iteration CNN weights, but the headline quality claim hinges on an unspecified evaluation reference that may be the same FDK target used in training. read the letter →

arxiv 2501.13128 v1 pith:VQSCJW2J submitted 2025-01-21 eess.IV cs.LG

Aniket Pramanik , Singanallur V. Venkatakrishnan , Obaidullah Rahman , Amirkoushyar Ziabari This is my paper · ORCID

classification eess.IVcs.LG

keywords ConvolutionalneuralnetworkCone-beamCTsparse-viewreconstructionhalf-quadraticsplittingplug-and-playpriorsdataconsistencyindustrialXCTdeeplearning

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 tries to establish that a hybrid reconstruction algorithm, which alternates a learnt CNN denoiser with a data-consistency step derived from cone-beam CT physics, can reconstruct high-quality volumes from sparse-view industrial scans (75 views instead of 1200) in about 30 seconds, with quality better than model-based iterative reconstruction (MBIR) and single-step deep learning baselines. The authors argue that this closes a practical gap: MBIR is too slow, single-step deep learning generalizes poorly, and deep unrolled networks are memory-hungry. They demonstrate the algorithm on the public Walnuts CBCT dataset, using only one scan for training and six for testing. If correct, the method makes sparse-view high-quality XCT practical for large industrial volumes.

What carries the argument

The central mechanism is a learnt half-quadratic splitting (HQS) scheme, in which the regularized least-squares problem is split into a Gaussian denoising subproblem and a data-consistency subproblem, solved alternately for a fixed K iterations. The denoising subproblem is approximated by a CNN D_theta, making it a learned proximal operator in the plug-and-play style, and the data-consistency subproblem is solved approximately by a fixed number of conjugate-gradient steps using the ASTRA forward operator. The paper's key design choice is training a separate CNN for each outer iteration (unshared weights) rather than one shared network, and doing so memory-efficiently with 2D patches extracted from the 3D volume. The regularization parameter beta balances denoising against data fidelity and can be adjusted at inference time, and for K=3 this alternation iteratively refines the reconstruction with the CG step anchoring the result to the measured projections.

What would settle it

Train the same algorithm with a dense-view MBIR reconstruction as ground truth (or on a numerical phantom with known density) instead of FDK-1200, and compare the resulting PSNR/SSIM against MBIR-75. If the advantage over MBIR disappears, the claim that the learned scheme is inherently superior is falsified; it was an artifact of the supervision target.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that replacing the proximal operator in half-quadratic splitting with a per-iteration convolutional neural network, and enforcing data consistency with a conjugate-gradient solve, yields reconstructions that outperform all the other methods on sparse-view Walnuts data. With K=3 outer iterations and unshared CNN weights, the method reaches PSNR/SSIM values above MBIR, U-Net, and a shared-weights variant, while taking 30 seconds per volume compared with MBIR's 1440 seconds. The paper also finds that unshared per-iteration weights matter: a shared-weight version over-smooths and degrades with iterations, whereas unshared weights sharpen edges and improve metrics. The method is memory-efficient because each iteration's CNN is trained separately on 2D patches, and the regularization parameter beta acts as a test-time knob for unseen imaging conditions.

Load-bearing premise

The load-bearing premise is that the dense-view FDK reconstruction used as the training target is a faithful, artifact-free proxy for the true object; if it contains systematic artifacts, the learned denoiser will preserve and amplify them, and the reported advantage over MBIR could be an artifact of supervision.

Editorial extensions

If this is right

Sparse-view XCT with 16x fewer projections becomes practical: reconstruction takes about 30 seconds per volume instead of about 1440 seconds for MBIR, enabling faster industrial scanning workflows.
The unshared-weights variant yields progressively sharper reconstructions across iterations, while shared weights over-smooth, so per-iteration training is a key contributor to quality.
The method generalizes to unseen sparse-view settings: at 50 views, a scenario not seen in training, it outperforms U-Net, suggesting better robustness to out-of-distribution measurement conditions.
Memory cost stays flat with respect to iterations (O(1)) rather than scaling with K as in deep unrolled end-to-end training, making the approach applicable to large industrial volumes.
The beta knob allows test-time trade-off between denoising and data consistency without retraining, giving users control over image quality when measurement conditions vary.

Reading between the lines

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

The reported superiority may partly derive from the choice of training target (dense-view FDK); if FDK-1200 contains systematic cone-beam artifacts, the CNN will reproduce them, so the comparison against MBIR could change if a different ground truth were used.
A natural test is to retrain the pipeline with MBIR-1200 as ground truth and compare; if the ranking flips, the current results reflect supervision rather than the iterative scheme itself.
The method still requires an accurate forward model A and its adjoint, so on industrial CBCT with large cone angles the ASTRA operator's accuracy is a hidden assumption that may limit generalization across scanner geometries.
One could extend the approach to 3D CNNs or learned denoisers with uncertainty estimates, but the memory-efficient 2D patch training would need rethinking for isotropic resolution.

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.

Desk Editor's Note

read the letter

The short version: this is a competent plug-and-play half-quadratic splitting paper with a practical memory-efficient twist, but the headline empirical claim is not yet supported because the paper never says what reference image the PSNR/SSIM numbers are computed against.

What is genuinely new: training each iteration's CNN separately with unshared weights on 2D patches, applied to large-scale CBCT. That is a sensible way to avoid the O(K) memory cost of deep unrolling, and the runtime table is useful: 30 seconds per volume versus 1440 for MBIR. The HQS math is standard and correctly presented.

The soft spot is the evaluation. The training loss in Eq. (9) uses the 1200-view FDK reconstruction as ground truth. The metrics in Fig. 3 are reported on zoomed patches, but the text never states what the reference image is. If that reference is the same FDK-1200 reconstruction, then the proposed method is being scored on how well it mimics its own training target, while MBIR is scored against a different objective. The ~0.5 dB margin over MBIR would then mean little. The authors need to state the reference explicitly and ideally evaluate against an independent ground truth, such as a phantom or a higher-quality reconstruction. Metrics on full volumes and error bars across the six walnuts would also help; right now we only see zoomed patches. No code is released, which makes it harder to check the numbers.

A second issue: the comparison omits deep-unrolled baselines that the paper itself cites (e.g., Lahiri et al.). Even if memory efficiency is the motivation, a comparison on the same Walnut data would contextualize the quality claims. The unqualified "state-of-the-art" claim in the conclusion overreaches on the basis of six walnuts.

The out-of-distribution test at 50 views is a good addition, but it inherits the same potential circularity if the reference is still FDK-1200.

These are addressable. The core idea has merit and the runtime advantage is real. As written, the empirical support for "outperforms all other methods" is not convincing. This paper deserves a serious referee, but the referee should push for a clarified evaluation protocol, full-volume metrics, and an independent reference. I would not cite the quality results until those are fixed, though I might cite the memory-efficient training idea.

Referee Report

3 major / 6 minor

Summary. The paper proposes a learnt half-quadratic splitting (HQS) algorithm for sparse-view industrial cone-beam CT reconstruction. The method alternates a 2D CNN denoiser, trained separately per iteration, with a conjugate-gradient data-consistency step, and uses unshared per-iteration weights. Experiments on the public Walnut CBCT dataset compare the proposed method against FDK, MBIR, a single-step U-Net, and a shared-weights variant, reporting PSNR/SSIM on zoomed patches and run-times. The paper claims the proposed method outperforms all compared methods and is more than an order of magnitude faster than MBIR.

Significance. If the empirical claims hold, the method offers a practical middle ground between computationally expensive MBIR and single-step deep learning: it retains a physics-based data-consistency block, yet requires few outer iterations and avoids the large memory footprint of end-to-end deep unrolling. The separate per-iteration training of 2D CNNs is a sensible strategy for large 3D volumes, and the use of a public benchmark is a strength. However, the central claim is currently supported by an evaluation whose reference image is not stated, and the comparison omits deep-unrolled baselines that the paper itself identifies as the relevant alternative class.

major comments (3)

[Section III-B, Fig. 3] The reference image used to compute PSNR/SSIM is never specified. In Eq. (9) the training target is the 1200-view FDK reconstruction xFDK; if the same xFDK is also the evaluation reference, the comparison is circular for the learnt component because the CNN is explicitly trained to output that image, while MBIR minimizes a different objective. If a different reference is used, it must be identified. The reported PSNR for FDK-1200 (about 10 dB) is implausibly low for a dense-view reconstruction against any sensible reference, which suggests either an error or an unusual reference; this must be corrected. The central claim of superior reconstruction quality requires evaluation against an independent reference (for example, a high-quality iterative reconstruction from all 1200 views, or a simulation with known ground truth) and full-volume metrics, not only zoomed patches.
[Section III-B and abstract] The comparison set contains only FDK, MBIR, U-Net, and the shared-weights variant. The introduction explicitly discusses deep-unrolled (DU) networks [7, 8, 9] as a competing class and motivates the proposed method as a memory-efficient alternative to them. The abstract and conclusion nevertheless claim that the proposed method "outperforms all the other methods" or "outperforms the state-of-the-art methods." At least one DU baseline (for example, FISTA-Net [7] or the method in [9]) should be included in the comparison, or the claim should be restricted to the methods actually evaluated.
[Section III-B] The empirical evidence consists of six test scans with a single PSNR/SSIM number per method and no error bars or statistical analysis. The reported advantages over MBIR are on the order of 0.2 to 0.5 dB, so without an assessment of variance it is difficult to judge whether these differences are significant. The authors should report error bars over the test volumes (and, if feasible, over multiple training runs) and should also report full-volume metrics, not only metrics on unspecified zoomed patches.

minor comments (6)

[Eqs. (4) and (7)] Equation (4) uses A^T while Eq. (7) uses A^H; the notation should be made consistent.
[Section III-A] The choice K = 3 is empirical and no convergence analysis is provided. Since the shared-weights variant degrades with iterations (Fig. 4), the paper should clarify that the method is a learned unrolled scheme and that no convergence to a fixed point of the HQS problem is claimed.
[Table I] The table lists FDK at 4 s and U-Net at 5 s, but the text says "U-Net is much faster compared to the other two" (referring to MBIR and the proposed method). FDK is also fast, so the wording should be adjusted to avoid confusion.
[General] The paper does not state whether code or trained models will be released; please add a reproducibility statement.
[Fig. 3] The description of the figure says PSNR/SSIM are reported for "zoomed patches," but the criteria for selecting those patches are not given. This should be specified so the reader can interpret the quantitative comparison.
[Abstract and Conclusion] The phrase "outperforms other methods" should be qualified by naming the specific comparison set and dataset, as the current wording implies a broader claim than the experiments support.

Circularity Check

1 steps flagged · score 6.0 of 10

Evaluation reference is the same FDK-1200 target used as training ground truth; the headline outperformance is fitted to that metric.

fitted input called prediction [Eq. (9) (training loss) and Section III-B / Fig. 3 (evaluation metrics)]
"The ground truth for training is an FDK reconstruction with 1200 views. ... PSNR in dB and SSIM values are reported for the zoomed patches."

The training loss in Eq. (9) minimizes MSE between the CNN output and the FDK-1200 reconstruction (xFDK). The paper's headline comparison (Fig. 3) reports PSNR/SSIM values without specifying any reference other than this same FDK-1200 ground truth. PSNR is a monotonic function of MSE, so the evaluation metric is exactly the training objective. The proposed method is therefore trained to maximize the score against which it is judged, while MBIR (which minimizes a different data-fidelity plus hand-designed regularizer objective) is not. Consequently, the claimed 'outperforms all the other methods' largely reflects the method's fit to the evaluation reference, rather than an independent measure of reconstruction fidelity relative to the true object.

full rationale

The core derivation of the HQS algorithm and the plug-and-play replacement of the proximal step by a CNN are textbook and self-contained. The central empirical claim, however, rests on PSNR/SSIM comparisons that appear to use the same 1200-view FDK reconstruction as both the training target (Eq. 9) and the evaluation reference (Fig. 3 caption never defines another reference). Since the CNN is explicitly trained to minimize MSE to FDK-1200, the reported PSNR against FDK-1200 is the training loss in log scale, making the superiority over MBIR statistically forced by construction. The runtime advantage (Table I) is a measured quantity and not circular, but the quality advantage that gives it meaning is tainted. No other circular steps were found: the PnP and pyMBIR self-citations are standard tools, not load-bearing premises, and no uniqueness theorem is imported.

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

The method leans on standard HQS and plug-and-play assumptions. The main load-bearing choice not justified by prior literature is the use of 1200-view FDK as ground truth. Beta and K are hand-chosen; no convergence guarantee is given for K=3. No invented physical entities are introduced.

free parameters (4)

beta (regularization parameter) = 5e-2
Empirically chosen and kept fixed across iterations (Section III-A). It balances the CNN denoiser and the data-consistency term, and the reported quality depends on this choice.
K (number of outer iterations) = 3
Chosen design parameter; performance is reported only for K=3 (Section III-A), and Fig. 4 shows quality changes with iteration count.
Number of CG iterations = 10
The data-consistency step in Eq. (7) is solved approximately with ten CG iterations (Section III-A).
CNN training hyperparameters = patch 256x256, batch 64, epochs 500, lr 1e-4, Adam, U-Net with 64 initial channels
All chosen by hand and central to reproducing the trained model; no ablation is provided.

assumptions (4)

domain assumption The cone-beam forward model A (ASTRA) is an accurate linear model of the measurement process.
Invoked in Eqs. (1), (4), (7); the data-consistency step and all comparisons assume ASTRA's projector is faithful.
domain assumption The 1200-view FDK reconstruction is a suitable ground truth for supervised training.
Used in Eq. (9); if FDK-1200 contains artifacts, the learned CNN inherits them, which may bias the comparison against MBIR and U-Net.
domain assumption A CNN trained to map sparse-view FDK images to dense-view FDK images approximates the proximal operator of the implicit regularizer Q in HQS.
Section II, Eqs. (5)-(6); this is the plug-and-play assumption on which the algorithm's interpretation rests.
ad hoc to paper A fixed small number of HQS iterations (K=3) is sufficient for the reported performance, without convergence analysis.
Section III-A; the method is not proven to converge and performance is reported only for K=3.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Learnt Half-Quadratic Splitting-Based Algorithm for Fast and High-Quality Industrial Cone-beam CT Reconstruction." pith.science (2026). https://pith.science/paper/VQSCJW2J

@misc{pith2026250113128,
  author       = {Pith},
  title        = {Pith review of: A Learnt Half-Quadratic Splitting-Based Algorithm for Fast and High-Quality Industrial Cone-beam CT Reconstruction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VQSCJW2J}},
  note         = {Machine review of arXiv:2501.13128}
}

read the original abstract

Industrial X-ray cone-beam CT (XCT) scanners are widely used for scientific imaging and non-destructive characterization. Industrial CBCT scanners use large detectors containing millions of pixels and the subsequent 3D reconstructions can be of the order of billions of voxels. In order to obtain high-quality reconstruction when using typical analytic algorithms, the scan involves collecting a large number of projections/views which results in large measurement times - limiting the utility of the technique. Model-based iterative reconstruction (MBIR) algorithms can produce high-quality reconstructions from fast sparse-view CT scans, but are computationally expensive and hence are avoided in practice. Single-step deep-learning (DL) based methods have demonstrated that it is possible to obtain fast and high-quality reconstructions from sparse-view data but they do not generalize well to out-of-distribution scenarios. In this work, we propose a half-quadratic splitting-based algorithm that uses convolutional neural networks (CNN) in order to obtain high-quality reconstructions from large sparse-view cone-beam CT (CBCT) measurements while overcoming the challenges with typical approaches. The algorithm alternates between the application of a CNN and a conjugate gradient (CG) step enforcing data-consistency (DC). The proposed method outperforms other methods on the publicly available Walnuts data-set.

Figures

Figures reproduced from arXiv: 2501.13128 by the authors.

**Figure 2.** Iterative rule of the proposed algorithm and its network architecture. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** Performance comparisons on Walnut Data. Walnut1 is the training data and the rest are testing data. Proposed-S-75 is the shared weights version of [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: Performance comparison of U-Net and the proposed method with [PITH_FULL_IMAGE:figures/full_fig_p003_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

16 extracted references · 12 canonical work pages

[7]

Fista-net: Learning a fast iterative shrinkage thresholding network for inverse problems in imaging,

J. Xiang, Y . Dong, and Y . Yang, “Fista-net: Learning a fast iterative shrinkage thresholding network for inverse problems in imaging,” IEEE Transactions on Medical Imaging , vol. 40, no. 5, pp. 1329–1339, 2021

2021
[9]

Sparse-view cone beam ct reconstruction using data-consistent supervised and adversarial learning from scarce training data,

A. Lahiri, G. Maliakal, M. L. Klasky, J. A. Fessler, and S. Ravis- hankar, “Sparse-view cone beam ct reconstruction using data-consistent supervised and adversarial learning from scarce training data,” IEEE Transactions on Computational Imaging , vol. 9, pp. 13–28, 2023

work page 2023
[1]

Practical cone-beam algorithm,

L. A. Feldkamp, L. C. Davis, and J. W. Kress, “Practical cone-beam algorithm,” Josa a, vol. 1, no. 6, pp. 612–619, 1984

1984
[2]

Fast model-based x-ray ct reconstruction using spatially nonhomogeneous icd optimization,

Z. Yu, J.-B. Thibault, C. A. Bouman, K. D. Sauer, and J. Hsieh, “Fast model-based x-ray ct reconstruction using spatially nonhomogeneous icd optimization,” IEEE Transactions on image processing , vol. 20, no. 1, pp. 161–175, 2010

work page 2010
[3]

A three- dimensional statistical approach to improved image quality for multislice helical ct,

J.-B. Thibault, K. D. Sauer, C. A. Bouman, and J. Hsieh, “A three- dimensional statistical approach to improved image quality for multislice helical ct,” Medical physics, vol. 34, no. 11, pp. 4526–4544, 2007

work page 2007
[4]

Deep con- volutional neural network for inverse problems in imaging,

K. H. Jin, M. T. McCann, E. Froustey, and M. Unser, “Deep con- volutional neural network for inverse problems in imaging,” IEEE Transactions on Image Processing, vol. 26, no. 9, pp. 4509–4522, 2017

work page 2017
[5]

Enabling rapid x-ray ct characterisation for additive manufacturing using cad models and deep learning-based reconstruc- tion,

A. Ziabari et al., “Enabling rapid x-ray ct characterisation for additive manufacturing using cad models and deep learning-based reconstruc- tion,” npj Computational Materials , vol. 9, no. 1, p. 91, 2023

work page 2023
[6]

Deep learning based workflow for accelerated industrial x-ray computed tomography,

O. Rahman et al. , “Deep learning based workflow for accelerated industrial x-ray computed tomography,” in 2023 IEEE International Conference on Image Processing (ICIP) . IEEE, 2023, pp. 2990–2994

work page 2023

Show all 16 references

[8]

Generalized deep iterative reconstruction for sparse-view ct imaging,

T. Su et al., “Generalized deep iterative reconstruction for sparse-view ct imaging,” Physics in Medicine & Biology , vol. 67, no. 2, p. 025005, 2022

2022
[10]

A cone-beam x-ray computed tomography data collection designed for machine learning,

H. Der Sarkissian et al., “A cone-beam x-ray computed tomography data collection designed for machine learning,” Scientific data, vol. 6, no. 1, p. 215, 2019

2019
[11]

Generalizing the optimized gradient method for smooth convex minimization,

D. Kim and J. A. Fessler, “Generalizing the optimized gradient method for smooth convex minimization,” SIAM Journal on Optimization , vol. 28, no. 2, pp. 1920–1950, 2018. [Online]. Available: https: //doi.org/10.1137/17M112124X

1920 doi
[12]

Plug-and-play priors for model based reconstruction,

S. V . Venkatakrishnan, C. A. Bouman, and B. Wohlberg, “Plug-and-play priors for model based reconstruction,” in 2013 IEEE Global Conference on Signal and Information Processing , 2013, pp. 945–948

2013
[13]

The astra toolbox: A platform for advanced algorithm development in electron tomography,

W. Van Aarle et al. , “The astra toolbox: A platform for advanced algorithm development in electron tomography,” Ultramicroscopy, vol. 157, pp. 35–47, 2015

2015
[14]

S. V . Venkatakrishnan, “pyMBIR,” 6 2019. [Online]. Available: https://www.osti.gov//servlets/purl/1550791

2019
[15]

U-net: Convolutional networks for biomedical image segmentation,

O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015 . Springer, 2015, pp. 234–241

2015
[16]

Fast and flexible x-ray tomography using the astra toolbox,

W. Van Aarle et al., “Fast and flexible x-ray tomography using the astra toolbox,” Optics express, vol. 24, no. 22, pp. 25 129–25 147, 2016

2016

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [7]

Fista-net: Learning a fast iterative shrinkage thresholding network for inverse problems in imaging,

J. Xiang, Y . Dong, and Y . Yang, “Fista-net: Learning a fast iterative shrinkage thresholding network for inverse problems in imaging,” IEEE Transactions on Medical Imaging , vol. 40, no. 5, pp. 1329–1339, 2021

2021

[2] [9]

Sparse-view cone beam ct reconstruction using data-consistent supervised and adversarial learning from scarce training data,

A. Lahiri, G. Maliakal, M. L. Klasky, J. A. Fessler, and S. Ravis- hankar, “Sparse-view cone beam ct reconstruction using data-consistent supervised and adversarial learning from scarce training data,” IEEE Transactions on Computational Imaging , vol. 9, pp. 13–28, 2023

work page 2023

[3] [1]

Practical cone-beam algorithm,

L. A. Feldkamp, L. C. Davis, and J. W. Kress, “Practical cone-beam algorithm,” Josa a, vol. 1, no. 6, pp. 612–619, 1984

1984

[4] [2]

Fast model-based x-ray ct reconstruction using spatially nonhomogeneous icd optimization,

Z. Yu, J.-B. Thibault, C. A. Bouman, K. D. Sauer, and J. Hsieh, “Fast model-based x-ray ct reconstruction using spatially nonhomogeneous icd optimization,” IEEE Transactions on image processing , vol. 20, no. 1, pp. 161–175, 2010

work page 2010

[5] [3]

A three- dimensional statistical approach to improved image quality for multislice helical ct,

J.-B. Thibault, K. D. Sauer, C. A. Bouman, and J. Hsieh, “A three- dimensional statistical approach to improved image quality for multislice helical ct,” Medical physics, vol. 34, no. 11, pp. 4526–4544, 2007

work page 2007

[6] [4]

Deep con- volutional neural network for inverse problems in imaging,

K. H. Jin, M. T. McCann, E. Froustey, and M. Unser, “Deep con- volutional neural network for inverse problems in imaging,” IEEE Transactions on Image Processing, vol. 26, no. 9, pp. 4509–4522, 2017

work page 2017

[7] [5]

Enabling rapid x-ray ct characterisation for additive manufacturing using cad models and deep learning-based reconstruc- tion,

A. Ziabari et al., “Enabling rapid x-ray ct characterisation for additive manufacturing using cad models and deep learning-based reconstruc- tion,” npj Computational Materials , vol. 9, no. 1, p. 91, 2023

work page 2023

[8] [6]

Deep learning based workflow for accelerated industrial x-ray computed tomography,

O. Rahman et al. , “Deep learning based workflow for accelerated industrial x-ray computed tomography,” in 2023 IEEE International Conference on Image Processing (ICIP) . IEEE, 2023, pp. 2990–2994

work page 2023