Pith. sign in

REVIEW 4 major objections 6 minor 32 references

GARD: Gamma-based Anatomical Restoration and Denoising for Retinal OCT

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read GARD claims that replacing the Gaussian noise assumption in diffusion-based denoising with a Gamma model of speckle, and guiding the reverse process with a non-locally filtered reference, substantially improves retinal OCT despeckling.

desk verdict A solid OCT denoising paper with a novel DDGM+NRFT combination, but the SOTA comparison is unfair and the real gain over vanilla DDGM is ~0.1 dB; worth refereeing after major fixes. read the letter →

arxiv 2509.10341 v1 pith:SIPBG23D submitted 2025-09-12 cs.CV

Botond Fazekas , Thomas Pinetz , Guilherme Aresta , Taha Emre , Hrvoje Bogunovic This is my paper · ORCID
classification cs.CV
keywords opticalcoherencetomographyspeckledenoisingdiffusionprobabilisticmodelsGammadistributionretinalimagingimagerestorationdeeplearningnoise-reducedfidelityterm
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

GARD is a diffusion-based denoiser for retinal OCT built on the premise that speckle noise in display-ready OCT is Gamma-distributed rather than Gaussian, and that a diffusion model which matches this statistic will remove noise without erasing fine anatomy. The paper claims that pairing this Gamma diffusion model with a Noise-Reduced Fidelity Term, a guide image obtained by non-local-means filtering the noisy scan, gives the best despeckling on a paired dataset of single-sweep and 30-frame-averaged B-scans, beating classical filters and state-of-the-art deep-learning baselines on PSNR, SSIM, and MSE. The authors argue that the fidelity term matters because enforcing consistency with the original noisy image, as prior diffusion denoisers do, tends to re-inject noise, whereas guiding with a less-noisy reference preserves structure while letting the model synthesize realistic high-frequency detail. If right, the method points to a practical path for getting clinically useful OCT quality from fewer acquisitions or lower-cost devices.

What carries the argument

The mechanism is the Gamma diffusion process from Eq. (1), where each forward step adds a centered Gamma-distributed variable with shape $k_t = \beta_t/(\alpha_t \theta_0^2)$ and scale $\theta_t = \sqrt{\bar{\alpha}_t} \theta_0$, chosen so that sums of independent Gamma variables with common scale remain Gamma-distributed, permitting closed-form sampling of $x_t$ from $x_0$. The reverse process in Eq. (2), with a U-Net trained to predict the noise component, is accelerated by setting $\sigma_t = 0$ (adapting the Denoising Diffusion Implicit Model approach for deterministic sampling), letting the denoiser start at $t=70$ and skip timesteps. The Noise-Reduced Fidelity Term (Eqs. 3-4) replaces the noisy input with its NLM-filtered version $\tilde{y} = \mathrm{NLM}(y)$ and, at each reverse step, solves a Newton optimization that balances the guide image against the diffusion output, retaining low-frequency anatomy while the diffusion fills in high-frequency detail.

What would settle it

A decisive check is to repeat the paired noisy/less-noisy evaluation on raw, linear-scale OCT B-scans before log or fourth-root compression, where speckle is known to be multiplicative. If GARD's advantage over the Gaussian DDPM shrinks or vanishes there, the additive-Gamma model is doing the load-bearing work; if the advantage persists, the NRFT guide, not the Gamma mechanism, drives the gains.

Watch

Extended reading notes

Core claim

The paper's central claim is that a Denoising Diffusion Gamma Model (DDGM), in which the forward noising process adds Gamma-distributed random variables rather than Gaussian ones, is a better statistical match to OCT speckle than standard DDPMs, and that this match translates into measurable denoising gains. The second claim is that the reverse diffusion should be steered not by the noisy input but by a non-local-means-filtered version of it, through the Noise-Reduced Fidelity Term; this prevents high-frequency noise from being reinforced. On a prospectively collected paired dataset, GARD reports the highest PSNR (28.25 dB), SSIM (0.58), and lowest MSE (103.95) among all compared methods, with Wilcoxon signed-rank tests marking every difference significant at p<0.01, and qualitative inspection showing crisper retinal-layer edges and better-preserved small reflective structures. The ablation results support the mechanism: the NRFT improves the Gamma DDGM but slightly degrades a Gaussian DDPM, which the authors read as evidence that the Gamma noise model and the noise-reduced guide work together.

Load-bearing premise

The load-bearing premise is that speckle in the display-ready, post-processed OCT images fed to the model is approximately additive and Gamma-distributed, so the diffusion forward and reverse processes match the real noise; if that statistical model is wrong for a given image domain, the learned reverse diffusion would not remove real speckle and the reported gains would rest on the NLM guide alone.

Editorial extensions

If this is right

  • GARD achieves the best PSNR, SSIM, and MSE on the paired noisy/less-noisy OCT dataset, with all differences significant at p<0.01, outperforming classical NLM and deep-learning baselines such as SCUNet, N2V2, and Speckle2Speckle.
  • The Noise-Reduced Fidelity Term improves the Gamma DDGM but slightly hurts a Gaussian DDPM, implying the noise-reduced guide is most effective when the diffusion model's noise statistics already match speckle.
  • A fidelity term that forces consistency with the original noisy image (CPDM) lowers performance, supporting the paper's argument that guiding with a less-noisy reference, not the noisy input, is the right design.
  • Because inference starts at t=70 and samples every 10th timestep via the Gamma-adapted DDIM, the method is substantially faster than a full 1000-step reverse diffusion, making clinical use more plausible.
  • Higher-quality images after denoising could enable faster acquisitions and cheaper OCT devices while preserving diagnostic detail, the application-level motivation.

Reading between the lines

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

  • If the Gamma-statistics premise transfers across devices, the $\theta_0$ hyperparameter and noise schedule would need per-device calibration; the paper's cross-vendor qualitative results hint at generalizability but do not quantify it, so a device-specific tuning study is the natural next test.
  • Because the NLM guide is a fixed preprocessing step, a learned or adaptively weighted guide could plausibly improve the fidelity term further; this is an extension the paper does not explore.
  • The same additive-Gamma-after-compression argument should apply to other coherent imaging modalities, such as ultrasound or other OCT systems, so a direct transfer test on log-compressed ultrasound speckle would reveal how much of the gain is generic rather than retinal-specific.
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. This paper presents GARD, a denoising method for retinal OCT B-scans that combines a Denoising Diffusion Gamma Model (DDGM) with a Noise-Reduced Fidelity Term (NRFT). The forward process in Eq. (1) adds zero-mean Gamma noise, and the reverse process in Eq. (2) follows Nachmani et al.; the NRFT replaces fidelity to the noisy input with fidelity to an NLM-filtered version (Eq. 3) via a proximal update (Eq. 4). DDIM-style deterministic sampling is used to accelerate inference. The method is trained on 2000 volumes from an investigational device and evaluated on 13 volumes with paired noisy (single sweep) and less-noisy (30-frame ART) B-scans from a commercial Spectralis device. Table 1 reports mean PSNR/SSIM/MSE for GARD, baselines (NLM, SCUNet, N2V2, Speckle2Speckle), and diffusion ablations, with Wilcoxon tests. GARD ranks first on all metrics, and qualitative results on cross-vendor data are shown. The authors claim significant improvement over state-of-the-art methods and sharper detail preservation.

Significance. The topic is relevant: OCT despeckling is an active clinical imaging problem, and paired noisy/less-noisy data with near-perfect registration are a valuable resource. The idea of adapting Gamma diffusion models to medical imaging is a novel extension, and the NRFT is a sensible mechanism to avoid noise reinforcement. The strengths of the paper include the paired-data evaluation design, the ablation across diffusion variants and fidelity terms, the use of a Wilcoxon test, and the release of source code. However, the significance is currently limited by the small single-device evaluation set and, more importantly, by the unfair comparison with external deep baselines that were not retrained on OCT. The actual margin over the strongest fairly trained baseline (vanilla DDGM) is 0.09 dB PSNR in Table 1, which is small and may not be practically meaningful. If the external baselines were retrained on the authors' training set, the state-of-the-art claim could plausibly disappear.

major comments (4)
  1. [Section 3 (Baselines), Table 1] The deep-learning baselines SCUNet, N2V2, and Speckle2Speckle are evaluated using publicly available pretrained weights without any retraining or fine-tuning on OCT data, as stated in Section 3. This is not an equal-footing comparison: SCUNet was trained on natural images, and it is unsurprising that a model trained on 2000 OCT volumes outperforms it off-the-shelf. The claim in Section 4 that GARD 'significantly outperforms state-of-the-art deep learning models' is therefore not established. Please either retrain these baselines on the same training set (or a comparable OCT corpus) and report the results, or explicitly rephrase the claim as a comparison against off-the-shelf pretrained models.
  2. [Section 4 vs Table 1] The effect sizes quoted in Section 4 are inconsistent with Table 1. The text states PSNR improvements of 0.31 dB over SCUNet, 0.34 dB over standard DDPM, and 0.23 dB over vanilla DDGM, but Table 1 gives GARD=28.25, SCUNet=28.10, DDPM=27.85, DDGM=28.16 (differences of 0.15, 0.40, and 0.09 dB, respectively). The 0.31 and 0.23 dB values do not match the table; please correct the text and discuss the actual margins, especially the small 0.09 dB gain over the vanilla DDGM, which is the only fairly trained diffusion baseline comparable to GARD.
  3. [Section 4 (Wilcoxon tests)] The statistical significance tests are reported as p<0.01 for all metrics and methods, but the paper does not specify the unit of analysis. If the 247 B-scans (13 volumes × 19 B-scans) are treated as independent samples, the test ignores within-volume correlation and overstates significance; if the unit is 13 volumes, the sample size is very small and the test's power is low. Please state the unit, report exact p-values, and, if B-scans are used, account for clustering (e.g., by volume-level averaging or a mixed-effects model).
  4. [Section 2 (Eq. 1) and Section 3 (Implementation)] The central modeling assumption is that display-ready OCT noise is approximately additive and Gamma-distributed, but this is not empirically validated. The hyperparameter θ0 is set to 0.1 because it 'resulted in a noise most similar to typical OCT noise' (Section 3), which is subjective. The paper should provide a quantitative comparison of the assumed Gamma distribution against measured residual noise statistics (e.g., on paired noisy/less-noisy B-scans), and also demonstrate that the reverse process removes real speckle beyond what the NRFT guide already achieves. Without such validation, the claimed advantage of the Gamma model over the Gaussian DDPM is only supported by a 0.09 dB PSNR difference in Table 1.
minor comments (6)
  1. [Section 2, Eq. (4)] The displayed optimization is garbled: 'z+e ˜y−z +µ' does not define a proper objective. Please rewrite the proximal update with explicit notation for the variables and the quadratic term.
  2. [Section 3, Datasets] For the quantitative evaluation set, specify the total number of B-scan pairs (13 volumes × 19 B-scans) and clarify whether the metrics in Table 1 are computed per B-scan and then averaged, or per volume.
  3. [Section 4 (statistical reporting)] The sentence 'GARD is significantly better for all metrics and methods with p<0.01' could be made more informative by reporting exact p-values or a supplementary table, and by stating whether any multiple-comparison correction was applied.
  4. [Section 2, Eq. (2)] The definition of \bar{g}_t and its shape parameter \bar{k}_t should be spelled out; it is only implicit that the sum of Gamma variables with the same scale is Gamma-distributed, and this property should be stated explicitly for the reader.
  5. [Figure 1] The column label 'Original Crop' is confusing because the actual crop is shown in the second column; adjust the caption or labels to match the content.
  6. [Section 3 and Table 1] The abbreviation for the fidelity term in [17] is written both as 'CPDM' and 'CDPM'; use one notation consistently throughout the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: GARD's core DDGM formulation is cited from external work [20], the NRFT guide is an NLM-filtered version of the noisy input (not of the evaluation target), and quantitative evaluation uses independently acquired ART-averaged B-scans that are never used to fit or define the method.

full rationale

The paper's derivation chain is self-contained against an external benchmark. The Gamma diffusion forward and reverse processes (Eqs. 1 and 2) are adopted from Nachmani et al. [20], an external citation, and the DDIM acceleration from Song et al. [27] is also external. No load-bearing step reduces to a self-citation by the present authors. The Noise-Reduced Fidelity Term (NRFT) uses a non-local means filtered version of the noisy input (Eq. 3) as a guide; this guide is a preprocessing of the input and is not derived from, nor does it coincide with, the evaluation target, which is a separately acquired 30-frame ART average. The evaluation therefore compares against an independent less-noisy reference, not against the method's own guide. Hyperparameters such as θ0, µ, and the starting timestep are reported as fixed choices, not fitted to the paired evaluation set. The only notable discrepancy is between the prose effect sizes (0.23 dB over DDGM, 0.31 dB over SCUNet) and the values implied by Table 1 (0.09 dB and 0.15 dB); this is a numerical reporting inconsistency, not circularity. Off-the-shelf pretrained baselines may weaken the fairness of the SOTA comparison, but that is a correctness/experimental-design concern, not a circular-derivation concern. No equation or claim is equivalent to its own input by construction, so the circularity score is 0.

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

The method relies on a small set of hyperparameters and several domain assumptions, but introduces no new physical entities. The most load-bearing assumption is the Gamma noise model in the transformed image domain, which is plausible but not empirically validated. The NLM guide and paired-data gold standard are additional domain assumptions.

free parameters (4)
  • theta0 (Gamma initial noise scale) = 0.1
    Chosen to make the synthetic Gamma noise resemble typical OCT noise; no sensitivity analysis or validation-based selection is reported.
  • mu (fidelity weight) = 10
    Set to enforce high consistency with the NLM-filtered image; no ablation over mu is presented.
  • Inference start time and step size = start at t=70, sample every 10th step
    Reduces the reverse process to 7 steps; the paper states this does not sacrifice quality but provides no quantitative comparison with the full reverse process.
  • NLM filter parameters = Not reported
    The non-local means pre-processing uses scikit-image defaults, but the specific patch size, distance, and smoothing parameters are not given, even though the fidelity guide depends on them.
assumptions (4)
  • domain assumption Speckle noise in the display-domain OCT is approximately additive and Gamma-distributed after dynamic range compression.
    This justifies the DDGM forward process (Eq. 1) and the reverse process (Eq. 2). If the real noise is not Gamma or not additive, the learned reverse diffusion will not match the actual degradation.
  • domain assumption NLM filtering preserves anatomical structures while reducing noise.
    The NRFT guides the reverse process toward the NLM output; if NLM smooths away fine retinal details, the fidelity term would enforce the loss of those details.
  • domain assumption The paired data (single-acquisition vs. 30-registered-average B-scans) represents a valid gold standard for denoising evaluation.
    The evaluation assumes the 30-average images are sufficiently clean and that the perfect registration makes pixel-wise comparison meaningful.
  • standard math The DDGM derivation and closed-form posterior from Nachmani et al. [20] are correct.
    The reverse process in Eq. (2) is taken directly from that paper and is not re-derived or verified here.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GARD: Gamma-based Anatomical Restoration and Denoising for Retinal OCT." pith.science (2026). https://pith.science/paper/SIPBG23D

@misc{pith2026250910341,
  author       = {Pith},
  title        = {Pith review of: GARD: Gamma-based Anatomical Restoration and Denoising for Retinal OCT},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SIPBG23D}},
  note         = {Machine review of arXiv:2509.10341}
}
read the original abstract

Optical Coherence Tomography (OCT) is a vital imaging modality for diagnosing and monitoring retinal diseases. However, OCT images are inherently degraded by speckle noise, which obscures fine details and hinders accurate interpretation. While numerous denoising methods exist, many struggle to balance noise reduction with the preservation of crucial anatomical structures. This paper introduces GARD (Gamma-based Anatomical Restoration and Denoising), a novel deep learning approach for OCT image despeckling that leverages the strengths of diffusion probabilistic models. Unlike conventional diffusion models that assume Gaussian noise, GARD employs a Denoising Diffusion Gamma Model to more accurately reflect the statistical properties of speckle. Furthermore, we introduce a Noise-Reduced Fidelity Term that utilizes a pre-processed, less-noisy image to guide the denoising process. This crucial addition prevents the reintroduction of high-frequency noise. We accelerate the inference process by adapting the Denoising Diffusion Implicit Model framework to our Gamma-based model. Experiments on a dataset with paired noisy and less-noisy OCT B-scans demonstrate that GARD significantly outperforms traditional denoising methods and state-of-the-art deep learning models in terms of PSNR, SSIM, and MSE. Qualitative results confirm that GARD produces sharper edges and better preserves fine anatomical details.

Figures

Figures reproduced from arXiv: 2509.10341 by the authors.

Figure 1
Figure 1. Qualitative denoising results on a OCT B-scan. Columns: Original noisy image with crop region indicated, Cropped noisy image, and the results for the best performing baseline model SCUNet, DDPM, GARD (Ours), and less-noisy reference target. GARD provides the sharpest edges and best detail preservation. Noisy input GARD Reference (Spectralis) Cirrus Cirrus Topcon [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Qualitative comparison of GARD denoising on OCT scans from different devices. Each row represents a different eye. Spectralis scans were acquired with ART and serve as less-noisy references [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 23 canonical work pages

  1. [1]

    Aum, J., hyun Kim, J., Jeong, J.: Effective speckle noise suppression in op- tical coherence tomography images using nonlocal means denoising filter with double gaussian anisotropic kernels. Appl. Opt.54(13), D43–D50 (May 2015). https://doi.org/10.1364/AO.54.000D43

    work page doi:10.1364/ao.54.000d43 2015

  2. [2]

    In: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05)

    Buades, A., Coll, B., Morel, J.M.: A non-local algorithm for image de- noising. In: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05). vol. 2, pp. 60–65 vol. 2 (Jun 2005). https://doi.org/10.1109/CVPR.2005.38, iSSN: 1063-6919

    doi:10.1109/cvpr.2005.38 2005

  3. [3]

    IEEE Transactions on Medical Imaging35(10), 2270–2279 (Oct 2016)

    Cheng, J., Tao, D., Quan, Y., Wong, D.W.K., Cheung, G.C.M., Akiba, M., Liu, J.: Speckle Reduction in 3D Optical Coherence Tomography of Retina by A-Scan Reconstruction. IEEE Transactions on Medical Imaging35(10), 2270–2279 (Oct 2016). https://doi.org/10.1109/TMI.2016.2556080

    arXiv 2016

  4. [4]

    Optics Communications 291, 461–469 (Mar 2013)

    Chong, B., Zhu, Y.K.: Speckle reduction in optical coherence tomography images of human finger skin by wavelet modified BM3D filter. Optics Communications 291, 461–469 (Mar 2013). https://doi.org/10.1016/j.optcom.2012.10.053

    work page doi:10.1016/j.optcom.2012.10.053 2013

  5. [5]

    IEEE Transactions on Medical Imaging42(4), 922– 934 (Apr 2023)

    Chung, H., Lee, E.S., Ye, J.C.: MR Image Denoising and Super-Resolution Using Regularized Reverse Diffusion. IEEE Transactions on Medical Imaging42(4), 922– 934 (Apr 2023). https://doi.org/10.1109/TMI.2022.3220681

    arXiv 2023

  6. [6]

    Roberts and Company Publishers (2007) 10 Fazekas et al

    Goodman, J.W.: Speckle phenomena in optics: theory and applications. Roberts and Company Publishers (2007) 10 Fazekas et al

    work page 2007

  7. [7]

    Göbl, R., Hennersperger, C., Navab, N.: Speckle2speckle: Unsupervised learning of ultrasound speckle filtering without clean data (2022)

    work page 2022

  8. [8]

    Archives of ophthalmology113(3), 325–332 (1995)

    Hee, M.R., Izatt, J.A., Swanson, E.A., Huang, D., Schuman, J.S., Lin, C.P., Pu- liafito, C.A., Fujimoto, J.G.: Optical coherence tomography of the human retina. Archives of ophthalmology113(3), 325–332 (1995)

    work page 1995

Show all 32 references
  1. [9]

    In: Ad- vances in Neural Information Processing Systems

    Ho, J., Jain, A., Abbeel, P.: Denoising Diffusion Probabilistic Models. In: Ad- vances in Neural Information Processing Systems. vol. 33, pp. 6840–6851. Curran Associates, Inc. (2020)

    2020

  2. [10]

    Science254(5035), 1178–1181 (1991)

    Huang, D., Swanson, E.A., Lin, C.P., Schuman, J.S., Stinson, W.G., Chang, W., Hee, M.R., Flotte, T., Gregory, K., Puliafito, C.A.: Optical coherence tomography. Science254(5035), 1178–1181 (1991)

    1991

  3. [11]

    In: Computer Vision – ECCV 2022 Workshops

    Höck, E., Buchholz, T.O., Brachmann, A., Jug, F., Freytag, A.: N2V2 - Fixing Noise2Void Checkerboard Artifacts with Modified Sampling Strate- gies and a Tweaked Network Architecture. In: Computer Vision – ECCV 2022 Workshops. pp. 503–518. Springer Nature Switzerland, Cham (202...

    2023 doi

  4. [12]

    IEEE Transactions on Medical Imaging34(5), 1042–1062 (2015)

    Kafieh, R., Rabbani, H., Selesnick, I.: Three dimensional data-driven multi scale atomic representation of optical coherence tomography. IEEE Transactions on Medical Imaging34(5), 1042–1062 (2015). https://doi.org/10.1109/TMI.2014.2374354

    2015

  5. [13]

    Karamata, B., Hassler, K., Laubscher, M., Lasser, T.: Speckle statistics in op- tical coherence tomography. J. Opt. Soc. Am. A22(4), 593–596 (Apr 2005). https://doi.org/10.1364/JOSAA.22.000593

    2005 doi

  6. [14]

    Journal of Biomedical Optics 21(7), 076008 (Jul 2016)

    Kopriva, I., Shi, F., Chen, X.: Enhanced low-rank + sparsity decomposition for speckle reduction in optical coherence tomography. Journal of Biomedical Optics 21(7), 076008 (Jul 2016). https://doi.org/10.1117/1.JBO.21.7.076008

    2016 doi

  7. [15]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (June 2019)

    Krull, A., Buchholz, T.O., Jug, F.: Noise2void - learning denoising from single noisy images. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (June 2019)

    2019

  8. [16]

    Progress in Retinal and Eye Research41, 26–43 (Jul 2014)

    Leitgeb, R.A., Werkmeister, R.M., Blatter, C., Schmetterer, L.: Doppler Optical Coherence Tomography. Progress in Retinal and Eye Research41, 26–43 (Jul 2014). https://doi.org/10.1016/j.preteyeres.2014.03.004

    2014 doi

  9. [17]

    In: MICCAI 2023

    Li, S., Higashita, R., Fu, H., Li, H., Niu, J., Liu, J.: Content-Preserving Diffusion Model for Unsupervised AS-OCT Image Despeckling. In: MICCAI 2023. pp. 660–

    2023

  10. [18]

    Nature Communications8(1), 15845 (Jun 2017)

    Liba, O., Lew, M.D., SoRelle, E.D., Dutta, R., Sen, D., Moshfeghi, D.M., Chu, S., de la Zerda, A.: Speckle-modulating optical coherence tomography in living mice and humans. Nature Communications8(1), 15845 (Jun 2017). https://doi.org/10.1038/ncomms15845

    2017 doi

  11. [19]

    Biomedical Optics Express9(11), 5129–5146 (Nov 2018)

    Ma, Y., Chen, X., Zhu, W., Cheng, X., Xiang, D., Shi, F.: Speckle noise reduction in optical coherence tomography images based on edge- sensitive cGAN. Biomedical Optics Express9(11), 5129–5146 (Nov 2018). https://doi.org/10.1364/BOE.9.005129

    2018 doi

  12. [20]

    https://doi.org/10.48550/arXiv.2110.05948, arXiv:2110.05948 [eess]

    Nachmani, E., Roman, R.S., Wolf, L.: Denoising Diffusion Gamma Models (Oct 2021). https://doi.org/10.48550/arXiv.2110.05948, arXiv:2110.05948 [eess]

    2021 doi arXiv

  13. [21]

    Springer (1999)

    Nocedal, J., Wright, S.J.: Numerical optimization. Springer (1999)

    1999

  14. [22]

    Journal of Biomedical Optics8(3), 565 – 569 (2003)

    Pircher, M., Götzinger, E., Leitgeb, R.A., Fercher, A.F., Hitzenberger, C.K.: Speckle reduction in optical coherence tomography by frequency compounding. Journal of Biomedical Optics8(3), 565 – 569 (2003). https://doi.org/10.1117/1.1578087 GARD: Gamma-based Anatomical Restorat...

    2003 doi

  15. [23]

    Optics Express17(2), 733–746 (Jan 2009)

    Puvanathasan, P., Bizheva, K.: Interval type-II fuzzy anisotropic diffusion algo- rithm for speckle noise reduction in optical coherence tomography images. Optics Express17(2), 733–746 (Jan 2009). https://doi.org/10.1364/OE.17.000733

    2009 doi

  16. [24]

    In: MICCAI 2015

    Ronneberger, O., Fischer, P., Brox, T.: U-Net: Convolutional networks for biomedical image segmentation. In: MICCAI 2015. pp. 234–241. Lecture Notes in Computer Science, Springer International Publishing, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28

    2015 doi

  17. [25]

    IEEE Transactions on Medical Imaging26(6), 761–771 (Jun 2007)

    Salinas, H.M., Fernandez, D.C.: Comparison of PDE-Based Nonlinear Diffusion Approaches for Image Enhancement and Denoising in Optical Coherence To- mography. IEEE Transactions on Medical Imaging26(6), 761–771 (Jun 2007). https://doi.org/10.1109/TMI.2006.887375

    2007

  18. [26]

    IEEE Jour- nal of Selected Topics in Quantum Electronics5(4), 1205–1215 (Jul 1999)

    Schmitt, J.: Optical coherence tomography (OCT): a review. IEEE Jour- nal of Selected Topics in Quantum Electronics5(4), 1205–1215 (Jul 1999). https://doi.org/10.1109/2944.796348

    1999

  19. [27]

    Song,J.,Meng,C.,Ermon,S.:Denoisingdiffusionimplicitmodels.In:International Conference on Learning Representations (2021)

    2021

  20. [28]

    Introduction, p

    WHO: World Population Ageing 2020 - Highlights, chap. Introduction, p. 3. World Health Organization (2020)

    2020

  21. [29]

    Zaki, F., Wang, Y., Su, H., Yuan, X., Liu, X.: Noise adaptive wavelet thresholding for speckle noise removal in optical coherence tomography. Biomed. Opt. Express 8(5), 2720–2731 (May 2017). https://doi.org/10.1364/BOE.8.002720

    2017 doi

  22. [30]

    Machine Intelligence Research20(6), 822–836 (Dec 2023)

    Zhang, K., Li, Y., Liang, J., Cao, J., Zhang, Y., Tang, H., Fan, D.P., Tim- ofte, R., Gool, L.V.: Practical Blind Image Denoising via Swin-Conv-UNet and Data Synthesis. Machine Intelligence Research20(6), 822–836 (Dec 2023). https://doi.org/10.1007/s11633-023-1466-0

    2023 doi

  23. [31]

    Journal of Biomedical Optics19(6), 066005 (2014)

    Zhang, X., Li, L., Zhu, F., Hou, W.W., Chen, X.: Spiking cortical model–based nonlocal means method for speckle reduction in optical coher- ence tomography images. Journal of Biomedical Optics19(6), 066005 (2014). https://doi.org/10.1117/1.JBO.19.6.066005

    2014 doi

  24. [670]

    Springer Nature Switzerland, Cham (2023)

    2023

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.