Speckle2Self: Self-Supervised Ultrasound Speckle Reduction Without Clean Data

Nassir Navab; Xuesong Li; Zhongliang Jiang

REVIEW 4 major objections 7 minor 16 references

Speckle2Self: Self-Supervised Ultrasound Speckle Reduction Without Clean Data

T0 review · 4 major / 7 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A single noisy ultrasound image is enough to train a speckle-removal network.

desk verdict A solid self-supervised despeckling paper whose central MSP assumption is partly violated by low-pass filtering; worth reviewing, but it needs AP-BSN/C-BSN baselines and a smoothing check. read the letter →

arxiv 2507.06828 v2 pith:DPO3X537 submitted 2025-07-09 eess.IV cs.AIcs.CV

Xuesong Li , Nassir Navab , Zhongliang Jiang This is my paper

classification eess.IVcs.AIcs.CV

keywords ultrasoundspecklereductionself-superviseddenoisingmulti-scaleperturbationlow-rankpriorsingle-imagecarotidB-modeimaging

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

Speckle in ultrasound is deterministic and tissue-dependent, so standard self-supervised denoising that assumes independent pixel noise cannot remove it, and clean references are physically unavailable. This paper claims that a simple multi-scale perturbation of a single noisy B-mode image, downsampling and then upsampling at three scales, creates several views that share the same anatomical structure while carrying different speckle patterns. A network trained to reconstruct each view while enforcing consistency among reconstructions implicitly separates the low-rank clean tissue from the sparse speckle, without any clean data or paired observations. The authors show the result outperforms classical filters and existing self-supervised methods on simulated and human carotid ultrasound, generalizes to unseen probe frequencies and devices, and improves downstream vessel segmentation. If true, this makes speckle reduction available wherever a single B-mode frame exists, including portable and low-cost scanners.

What carries the argument

The load-bearing mechanism is the multi-scale perturbation (MSP): controlled downsampling and upsampling of the input at scale factors 1.0, 0.5, and 0.25, which is assumed to leave the clean tissue structure X unchanged while reshuffling the speckle pattern. Around this, the framework is a multi-encoder network, one encoder per scale sharing a single decoder, trained with a reconstruction loss (MSE to each perturbed view) and a consistency loss (L1 between all pairs of reconstructed outputs). The low-rank prior on clean tissue, the assumption that anatomy lies on a low-rank subspace while speckle is sparse or high-rank, is what gives the consistency loss its meaning: the network is rewarded for outputs that are common across scales and penalized for outputs that follow scale-specific speckle.

What would settle it

On a simulated phantom with a known clean image X and sharp point scatterers, run Speckle2Self and compare the output to X and to X passed through the same downsampling-upsampling at scale 0.25; if the output is closer (higher SSIM) to the smoothed version than to X, the assumption that MSP preserves structure fails and the method is doing low-pass estimation rather than true despeckling.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that ultrasound speckle can be removed from a single noisy image by treating the clean anatomy as a low-rank signal and the speckle as sparse, scale-dependent variation. The multi-scale perturbation (MSP) operation, with scale factors 1.0, 0.5, and 0.25, is claimed to preserve the shared anatomical structure X while producing distinct speckle realizations N(k) for each scale (Eq. 7). Three scale-specific encoders and a shared decoder are trained with an MSE reconstruction loss toward each perturbed input plus an L1 consistency loss between all pairs of reconstructions. The consistency loss forces the network to find what is invariant across scales, and the authors report that MSE reconstruction plus L1 consistency is the only loss combination that removes speckle, while symmetric combinations blur structure or retain speckle. The outcome is a despeckled image that, in their experiments, beats established filters (SRAD, NLM, BM3D, OBNLM) and prior self-supervised methods (N2N, N2V, Noise2Self, Neighbor2Neighbor, DIP, ZS-N2N) on SSIM, LPIPS, and homogeneity, with only a supervised method trained on clean targets scoring higher.

Load-bearing premise

The whole training signal rests on the premise that downsampling and upsampling changes only the speckle while leaving the true anatomy untouched, but downsampling also removes fine tissue detail, so the shared structure is really a smoothed image, which likely explains the reported boundary blurring.

Editorial extensions

If this is right

Speckle reduction no longer requires clean images, paired noisy observations, or plane-wave hardware; one B-mode frame suffices for training, with about 7 ms per image at inference.
Because no tissue-specific noise model is assumed, the same trained network transfers to unseen central frequencies (3.75 to 9.38 MHz) and to images from two unseen commercial scanners, with zero-shot use and small fine-tuning gains.
Reducing speckle improves downstream analysis: on carotid artery segmentation with a rule-based flood-fill segmenter, AUC rose from 0.589 to 0.816 and IoU from 0.175 to 0.497 after Speckle2Self denoising.
The method is architecture-light (about three million parameters) and operates on standard B-mode output, so it can be deployed on portable and point-of-care systems where clean data do not exist.

Reading between the lines

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

If cross-scale agreement is the right training signal, the same perturbation idea could be tested on other deterministic, spatially correlated image degradations, such as CT metal artifacts or MRI ghosting, where independent-noise self-supervision fails.
The reported boundary blurring suggests the method may be estimating a low-pass-filtered version of the anatomy rather than the true X; a direct test would compare denoised output against both clean X and its downsampled-upsampled version on synthetic data.
Because speckle is itself a deterministic signal used in elastography and motion tracking, the network's ability to separate speckle from tissue could be repurposed to extract speckle maps, not only to discard them.
A cheap diagnostic for the mechanism is replacing the three-scale perturbation with two scales plus stronger perturbation to see whether performance tracks the degree of speckle decorrelation.

Share X Bluesky LinkedIn Reddit HN

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

Colleague, this one is worth a look if you track self-supervised denoising for medical imaging. The core idea: apply downsampling-upsampling at three scales to a single noisy B-mode ultrasound image, train a multi-encoder/shared-decoder network to reconstruct each perturbed input, and enforce L1 consistency among outputs. The claim is that the shared component across scales is the clean anatomy while speckle changes, so the network learns to suppress speckle without clean data. What's actually new is the domain-specific adaptation of downsampling-based self-supervision to deterministic, tissue-dependent ultrasound speckle, plus the three-encoder architecture and a solid set of experiments on synthetic and real carotid images.

Credit where due: the paper is transparent about its limitations—it admits the low-rank prior is an approximation and that MSP causes slight boundary blurring. The ablation study is thoughtful: it tests loss configurations, interpolation schemes, scale combinations, and compares MSP against low-pass filters as input perturbations. The cross-device generalization and downstream segmentation experiments are a plus. The synthetic results consistently beat classical filters and older self-supervised methods on SSIM, LPIPS, and homogeneity, even if PSNR is not the best.

Soft spots, in order of importance. First, the most relevant baselines are missing: AP-BSN and C-BSN, which also train on downsampled image pairs, are cited but not evaluated. Without them, the novelty claim is under-supported. Second, the load-bearing assumption in Eq. (7) is not strictly true. Downsampling by 0.25 and upsampling back removes high-frequency tissue information, so the cross-scale shared structure is a low-pass filtered version of X, not X itself. The consistency loss can therefore be minimized by outputting a smoothed image. The paper acknowledges 'slight blurring along tissue boundaries,' but the real-data evaluation uses GLCM homogeneity, a metric that explicitly rewards smoothness, and there are no error bars or significance tests. On synthetic data, the wins are exactly on structure/perceptual metrics, not PSNR, which is consistent with a smoothing effect. This doesn't sink the paper—the method still outperforms BM3D and NLM on SSIM/LPIPS, so it's not just Gaussian blur—but the interpretation that MSP isolates 'sparse noise' needs more support. A simple low-pass baseline (e.g., Gaussian filter with tuned sigma) on the same synthetic data would clarify how much of the gain is smoothing. Third, the reported numbers are single runs; some error bars and a significance test would help.

Who this is for: researchers working on self-supervised denoising for ultrasound or other structured noise. It is a solid contribution that a serious editor should send to review, with the expectation that the baselines and the smoothing analysis get addressed.

Referee Report

4 major / 7 minor

Summary. The paper proposes Speckle2Self, a self-supervised method for ultrasound speckle reduction that requires only single noisy B-mode images. The method applies a multi-scale perturbation (MSP) operation, implemented as downsampling-upsampling at scales {1.0, 0.5, 0.25}, to create three variants of the input; it assumes each variant can be written as X + N^(k) with shared clean structure X and scale-specific speckle N^(k), and trains three encoders plus a shared decoder with an MSE reconstruction loss and an L1 cross-scale consistency loss, motivated by low-rank/sparse decomposition. The paper evaluates the method on three synthetic MUST datasets, in-vivo carotid images from a Clarius scanner, cross-configuration and cross-device generalization, a loss/interpolation/scale ablation, and a downstream carotid segmentation experiment.

Significance. The proposed setting is practically important: ultrasound speckle is tissue-dependent and deterministic, so N2N-style paired observations and blind-spot assumptions are often inapplicable, and a method that works from single images would fill a real gap. The paper's strengths are its extensive comparison set, the use of a realistic simulator, the cross-device generalization attempt, the ablation study, and the reported 7 ms inference time. The central weakness is that the MSP assumption has not been separated from low-pass smoothing: the metrics on which Speckle2Self wins are substantially the ones improved by blur, and the in-vivo evidence is a single smoothing-favoring metric. If the authors can provide evidence that the method preserves high-frequency tissue structure while suppressing speckle, this would be a valuable contribution to the field.

major comments (4)

[Section 3.3, Eq. (7)] The decomposition P_k(Y) = X + N^(k) is the load-bearing assumption, but downsampling-upsampling does not leave X unchanged: rescaling by 0.25 and interpolating back removes high-frequency tissue signal, so the component shared across scales is a low-pass filtered version of X, not X itself. With the loss in Eq. (11), the network can minimize both the MSE reconstruction term and the L1 consistency term by emitting a smoothed image that sits near all three filtered views; MSE is dominated by low-frequency fidelity. The Discussion's admission of 'slight blurring along tissue boundaries' is therefore not a cosmetic limitation but direct evidence for this mechanism, and the pattern in Table 4 (PSNR not best on S-I/S-II while SSIM/LPIPS/homogeneity improve) is consistent with low-pass filtering. Please provide a test that disentangles speckle suppression from smoothing, e.g., quantify P_k(X) - X on the synthetic clean targets, compare with a model trained on LPF perturbations of matched variation, or report a high-frequency/edge-preservation metric.
[Section 4.4.4, Table 4 (Carotid column)] The in-vivo evaluation rests on a single no-reference metric, GLCM homogeneity, which explicitly rewards local smoothness; no error bars, confidence intervals, or paired statistical tests are reported for the 104 test images. Given the smoothing concern above, this single metric cannot support the statement that Speckle2Self 'consistently outperforms' other methods on real data. Please report per-image distributions or mean +/- std, add a structure-sensitive no-reference metric (e.g., edge/gradient preservation), and ideally a blinded clinician preference study.
[Sections 4.4.1 and 4.6.2] The baseline set omits the most relevant self-supervised ultrasound methods, including the deep ultrasound denoising approach already cited as Goudarzi and Rivaz (2023) and the AP-BSN/C-BSN variants (Lee et al. 2022b; Jang et al. 2023) designed for spatially correlated noise; the plane-wave-based N2N variants cited in the Introduction are also not compared. The cross-device evaluation in Section 4.6.2 is qualitative only, with no metric reported on the Siemens and Cephasonics frames before or after fine-tuning. Without these quantitative comparisons, the claims of consistent superiority over self-supervised methods and strong cross-device generalization are not fully supported.
[Section 4.3] The SVD singular-value decay analysis shows that Speckle2Self outputs have faster singular-value decay, but any low-pass filter produces this effect; without a control for detail preservation, the experiment does not demonstrate that the method preserves tissue structure while removing speckle. On the synthetic data where ground truth exists, please complement this analysis with a quantitative measure of how much of the clean image's high-frequency structure is retained.

minor comments (7)

[Section 2.4] The heading 'Nosie2Void' contains a typo and should read 'Noise2Void'.
[Section 5] The heading 'Discusssion' contains a typo and should read 'Discussion'.
[Section 4.7] The text refers to the 'Folld fill algorithm'; this should be 'Flood Fill algorithm'.
[Section 4.5.2] In the sentence describing Figure 9, 'the proposed MPS' should be 'the proposed MSP'.
[Section 6] The sentence 'This denied result will enhance image consistency and thereby boost the development of reproducible computer-aided diagnosis' is malformed and unclear; it should be rewritten.
[Section 4.4.1] The reference for N2S(*) is given as 'Krull et al. (2019)', but Noise2Self should cite Batson and Royer (2019).
[Table 4] All synthetic and carotid metrics are reported as point estimates without standard deviations or confidence intervals, so the statistical significance of the performance gaps cannot be assessed.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the method is trained from single noisy images and validated on independent simulated ground truth; the low-rank assumption is a stated prior, not a fitted result.

full rationale

The derivation chain is self-contained and non-circular. The training signal is generated entirely from single noisy inputs via the MSP operation (Eq. 7), and the objective (Eq. 11) is a sum of reconstruction and cross-scale consistency losses with no fitted parameter tied to the reported metrics. Synthetic evaluation uses independently simulated clean ground truth (Sec. 4.2.1), and the method is benchmarked against external methods; there is no fitted input renamed as a prediction. The low-rank assumption (Sec. 3.2–3.3) is a stated prior, not a consequence of the data or of a self-citation. The Discussion's admission of 'slight blurring along tissue boundaries' is a correctness and assumption limitation, not circular reasoning. Self-citations to prior robotic-ultrasound work are contextual and not load-bearing; no central claim reduces to an author's prior result. The frequency-domain analysis in Sec. 3.3 is weak supporting evidence because low-frequency content is preserved by construction under downsampling, but this is not the core derivation and does not make the reported results equivalent to the inputs.

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

The method adds no free physical constants but relies on several modeling choices (scale factors, loss weighting, interpolation) and on the low-rank and separability assumptions for US speckle. No new physical entities are introduced.

free parameters (5)

MSP scale factors = s_k in {1.0, 0.5, 0.25}
Chosen to create views with sufficiently different speckle patterns; no optimality criterion given in Sec. 3.3.
Number of MSP scales = K = 3
Ablation shows three scales (H+M+L) outperforms two scales, but this was not systematically optimized beyond that configuration.
Loss weights = 1.0 for both L_rec and L_con
Eq. 11 sums the two losses with equal weight; no weight tuning is reported.
Interpolation method = bilinear (default)
Table 5 shows bilinear and bicubic are competitive; bilinear is chosen even though bicubic sometimes performs better on S-III.
Training hyperparameters = lr=0.001, batch=16, 3000 epochs
Standard Adam settings, not justified by sensitivity analysis.

assumptions (5)

domain assumption RF signal is multiplicative: Y_RF = X_RF * N_RF (Eq. 1).
Standard model of US speckle from coherent wave interference, cited to Szabo (2013).
domain assumption After log compression, speckle becomes approximately additive in B-mode (Eq. 2).
Standard approximation used by Coupé et al. (2009); the paper relies on this to justify denoising in B-mode.
ad hoc to paper Clean tissue signal X is low-rank; speckle N is sparse or high-rank.
Assumed in Sec. 3.3 without empirical validation for US images; acknowledged in Discussion as 'an approximation rather than a physically rigorous model'.
ad hoc to paper The MSP operation produces variants with identical X and different N^(k) (Eq. 7).
Core premise of the method; downsampling-upsampling changes speckle but also blurs high-frequency structure, so strict equality of X across scales is only approximate.
domain assumption Speckle in B-mode has non-zero mean and is deterministic, making N2N inapplicable.
Used to motivate the method and supported by US literature (Prince & Links 2006, Christensen et al. 2024).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Speckle2Self: Self-Supervised Ultrasound Speckle Reduction Without Clean Data." pith.science (2026). https://pith.science/paper/DPO3X537

@misc{pith2026250706828,
  author       = {Pith},
  title        = {Pith review of: Speckle2Self: Self-Supervised Ultrasound Speckle Reduction Without Clean Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DPO3X537}},
  note         = {Machine review of arXiv:2507.06828}
}

read the original abstract

Image denoising is a fundamental task in computer vision, particularly in medical ultrasound (US) imaging, where speckle noise significantly degrades image quality. Although recent advancements in deep neural networks have led to substantial improvements in denoising for natural images, these methods cannot be directly applied to US speckle noise, as it is not purely random. Instead, US speckle arises from complex wave interference within the body microstructure, making it tissue-dependent. This dependency means that obtaining two independent noisy observations of the same scene, as required by pioneering Noise2Noise, is not feasible. Additionally, blind-spot networks also cannot handle US speckle noise due to its high spatial dependency. To address this challenge, we introduce Speckle2Self, a novel self-supervised algorithm for speckle reduction using only single noisy observations. The key insight is that applying a multi-scale perturbation (MSP) operation introduces tissue-dependent variations in the speckle pattern across different scales, while preserving the shared anatomical structure. This enables effective speckle suppression by modeling the clean image as a low-rank signal and isolating the sparse noise component. To demonstrate its effectiveness, Speckle2Self is comprehensively compared with conventional filter-based denoising algorithms and SOTA learning-based methods, using both realistic simulated US images and human carotid US images. Additionally, data from multiple US machines are employed to evaluate model generalization and adaptability to images from unseen domains. Project page: https://noseefood.github.io/us-speckle2self/

Figures

Figures reproduced from arXiv: 2507.06828 by the authors.

**Figure 2.** Intuitive visualization of the image quality degradation caused by [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Frequency-domain analysis of image variants generated by MSP operation. Low-frequency components show high cross-scale correlation, indicating shared low-rank tissue structure, while high-frequency components vary significantly due to scale-induced speckle differences. 3.3. Implicit Low-Rank Approximation for US Speckle Classical low-rank approximation methods aim to explicitly decompose an image into low-rank and… view at source ↗

Figures from the paper (9 more)

**Figure 4.** Figure 4: Overview of our proposed despeckling framework. (a) Overall training process. (b) Inference using the trained despeckling model. The network generates [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: SVD-based singular value spectra of two image patches (top: ho [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

**Figure 6.** Figure 6: Despeckling results and average SSIM values on Synthetic S-II dataset. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

**Figure 7.** Figure 7: Despeckling results and average homogeneity values on carotid US dataset acquired from volunteers. [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]

**Figure 8.** Figure 8: Visual comparison of different loss configurations for training the proposed Speckle2Self network. (a) Clean ground truth. (b) Noisy input. (c) Our method (MSE + L1) achieves the best trade-off between speckle suppression and structural fidelity. (d)–(e) Homogeneous l…

**Figure 9.** Figure 9: Denoising performance comparison between low-pass filters and the proposed MSP operation as preprocessing. (a) Noisy input image. (b) and (c) Correlation matrices computed among the high-frequency component of three images generated using Gaussian smoothing filters wit…

**Figure 11.** Figure 11: Qualitative comparison of denoising performance on images ob [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]

**Figure 10.** Figure 10: Cross-configuration testing results show that higher central frequen [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]

**Figure 12.** Figure 12: Comparison of various speckle reduction methods in enhancing carotid ultrasound segmentation accuracy: Blue curves indicate the ground truth [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

16 extracted references · 11 canonical work pages

[5]

arXiv preprint arXiv:2504.00264

Diffdenoise: self-supervised medical image denoising with condi- tional diffusion models. arXiv preprint arXiv:2504.00264 . Frost, V .S., Stiles, J.A., Shanmugan, K.S., Holtzman, J.C.,

arXiv
[8]

Motion magnification in robotic sonography: Enabling pulsation-aware artery segmentation, in: 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE. pp. 6565–6570. Huang, D., Li, C., Karlas, A., Chu, X., Au, K.S., Navab, N., Jiang, Z.,

work page 2023
[10]

Speckle- constrained filtering of ultrasound images, in: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), IEEE. pp. 547–552. Krissian, K., Westin, C.F., Kikinis, R., V osburgh, K.G.,

work page 2005
[11]

Learning Sparse and Low-Rank Priors for Image Recovery via Iterative Reweighted Least Squares Minimization

Learning sparse and low-rank priors for image recovery via iterative reweighted least squares minimization. arXiv preprint arXiv:2304.10536 . Lehtinen, J.,

work page Pith review arXiv
[12]

arXiv preprint arXiv:1803.04189

Noise2noise: Learning image restoration without clean data. arXiv preprint arXiv:1803.04189 . Li, J., Su, T., Zhao, B., Lv, F., Wang, Q., Navab, N., Hu, Y ., Jiang, Z.,

arXiv
[15]

Cross-modal attention for mri and ultrasound volume regis- tration, in: Medical Image Computing and Computer Assisted Intervention– MICCAI 2021: 24th International Conference, Strasbourg, France, Septem- ber 27–October 1, 2021, Proceedings, Part IV 24, Springer. pp. 66–75. Stevens, T.S., Meral, F.C., Yu, J., Apostolakis, I.Z., Robert, J.L., Van Sloun, R.J.,

work page 2021
[1973]

IEEE Transactions on Systems, Man, and Cybernetics SMC- 3, 610–621

Textural features for image classification. IEEE Transactions on Systems, Man, and Cybernetics SMC- 3, 610–621. doi: 10.1109/TSMC.1973.4309314. He, K., Zhang, X., Ren, S., Sun, J.,

arXiv 1973
[2005]

A non-local algorithm for image denoising, in: 2005 IEEE computer society conference on computer vision and pattern recognition (CVPR’05), Ieee. pp. 60–65. Calis, M., Mischi, M., van der Veen, A.J., Hunyadi, B.,

work page 2005

Show all 16 references

[2007]

IEEE Transactions on image processing 16, 2080–2095

Image denoising by sparse 3-d transform-domain collaborative filtering. IEEE Transactions on image processing 16, 2080–2095. Dantas, R.G., Costa, E.T., Leeman, S.,
[2010]

Signal Processing 90, 1963–1975

Ramp preserving perona– malik model. Signal Processing 90, 1963–1975. Chen, Y ., Fang, B., Li, H., Huang, L., Luo, J.,

1963
[2012]

Practical low- rank matrix approximation under robust l 1-norm, in: 2012 IEEE Confer- ence on Computer Vision and Pattern Recognition, IEEE. pp. 1410–1417. Zhou, Y ., Bi, Y ., Tong, W., Wang, W., Navab, N., Jiang, Z.,

2012
[2018]

BioMed research international 2018, 5137904

Machine learning in ultrasound computer- aided diagnostic systems: a survey. BioMed research international 2018, 5137904. Huang, T., Li, S., Jia, X., Lu, H., Liu, J.,

2018
[2021]

A bias-reducing loss function for ct image denoising, in: ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE. pp. 1175–1179. Nyrnes, S.A., Fadnes, S., Wigen, M.S., Mertens, L., Lovstakken, L.,

2021
[2023]

Deep ultrasound denoising without clean data, in: Medical Imaging 2023: Ultrasonic Imaging and Tomography, SPIE. pp. 131–136. Gu, S., Li, Y ., Gool, L.V ., Timofte, R.,

2023
[2024]

arXiv preprint arXiv:2408.00714

Sam 2: Segment anything in images and videos. arXiv preprint arXiv:2408.00714 . Ronneberger, O., Fischer, P., Brox, T.,

arXiv
[2025]

arXiv preprint arXiv:2506.21499

Lightweight physics- informed zero-shot ultrasound plane wave denoising. arXiv preprint arXiv:2506.21499 . Batson, J., Royer, L.,

arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [5]

arXiv preprint arXiv:2504.00264

Diffdenoise: self-supervised medical image denoising with condi- tional diffusion models. arXiv preprint arXiv:2504.00264 . Frost, V .S., Stiles, J.A., Shanmugan, K.S., Holtzman, J.C.,

arXiv

[2] [8]

Motion magnification in robotic sonography: Enabling pulsation-aware artery segmentation, in: 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE. pp. 6565–6570. Huang, D., Li, C., Karlas, A., Chu, X., Au, K.S., Navab, N., Jiang, Z.,

work page 2023

[3] [10]

Speckle- constrained filtering of ultrasound images, in: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), IEEE. pp. 547–552. Krissian, K., Westin, C.F., Kikinis, R., V osburgh, K.G.,

work page 2005

[4] [11]

Learning Sparse and Low-Rank Priors for Image Recovery via Iterative Reweighted Least Squares Minimization

Learning sparse and low-rank priors for image recovery via iterative reweighted least squares minimization. arXiv preprint arXiv:2304.10536 . Lehtinen, J.,

work page Pith review arXiv

[5] [12]

arXiv preprint arXiv:1803.04189

Noise2noise: Learning image restoration without clean data. arXiv preprint arXiv:1803.04189 . Li, J., Su, T., Zhao, B., Lv, F., Wang, Q., Navab, N., Hu, Y ., Jiang, Z.,

arXiv

[6] [15]

Cross-modal attention for mri and ultrasound volume regis- tration, in: Medical Image Computing and Computer Assisted Intervention– MICCAI 2021: 24th International Conference, Strasbourg, France, Septem- ber 27–October 1, 2021, Proceedings, Part IV 24, Springer. pp. 66–75. Stevens, T.S., Meral, F.C., Yu, J., Apostolakis, I.Z., Robert, J.L., Van Sloun, R.J.,

work page 2021

[7] [1973]

IEEE Transactions on Systems, Man, and Cybernetics SMC- 3, 610–621

Textural features for image classification. IEEE Transactions on Systems, Man, and Cybernetics SMC- 3, 610–621. doi: 10.1109/TSMC.1973.4309314. He, K., Zhang, X., Ren, S., Sun, J.,

arXiv 1973

[8] [2005]

A non-local algorithm for image denoising, in: 2005 IEEE computer society conference on computer vision and pattern recognition (CVPR’05), Ieee. pp. 60–65. Calis, M., Mischi, M., van der Veen, A.J., Hunyadi, B.,

work page 2005