VAMOS-OCTA: Vessel-Aware Multi-Axis Orthogonal Supervision for Inpainting Motion-Corrupted OCT Angiography Volumes

Ehsan Khodapanah Aghdam; Han Liu; Hao Li; Ipek Oguz; Jacob Watson; Nick DiSanto; Yuankai K. Tao

REVIEW 4 major objections 5 minor 25 references

VAMOS-OCTA: Vessel-Aware Multi-Axis Orthogonal Supervision for Inpainting Motion-Corrupted OCT Angiography Volumes

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read Adding axial and lateral projection-consistency losses to a vessel-weighted reconstruction loss lets a 2.5D U-Net inpaint motion-corrupted OCTA B-scans, delivering sharp capillaries, restored vessel continuity, and clean en face projections

desk verdict A fairly honest incremental paper: a new loss recipe for OCTA B-scan inpainting that shows real gains on perceptual and MIP metrics, but the central claim of 'consistent outperformance' is undercut by the paper's own PSNR tables and by unvalidated synthetic-to-real transfer. read the letter →

arxiv 2602.00995 v1 pith:Y3XKKOPZ submitted 2026-02-01 cs.CV

Nick DiSanto , Ehsan Khodapanah Aghdam , Han Liu , Jacob Watson , Yuankai K. Tao , Hao Li , Ipek Oguz This is my paper · ORCID

classification cs.CV

keywords OCTAinpaintinghandheldretinalimagingmotionartifactcorrectionvessel-awarelossmulti-axisprojectionsupervision2.5DU-NetenfaceMIPrestorationmedicalimage

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

The paper aims to show that motion-related blind spots in handheld OCTA volumes—entire B-scans lost to sudden motion—can be filled in by a 2.5D U-Net trained with a composite loss that watches not only the slices themselves but also their one-dimensional projections along depth and across slices. The proposed VAMOS loss combines a vessel-weighted MSE with axial and lateral maximum/average intensity projection losses. The authors argue that these orthogonal projection constraints force the network to keep vessels continuous in 3D without building an explicit 3D model. If the claim holds, clinicians using handheld OCTA on children or uncooperative patients get sharper cross-sectional images and artifact-free en face maps from the same acquisition, and slice-based 3D imaging generally gains a simple loss-level tool for volumetric consistency.

What carries the argument

The load-bearing object is the VAMOS loss: L = wMSE + λ_proj (L_ax_MIP + L_lat_MIP + L_ax_AIP + L_lat_AIP). The wMSE weights each pixel by a combination of the target intensity and the predicted intensity raised to 1/3 (with constants 100 and 0.5), so bright vessel pixels dominate and hallucinated bright spots are penalized. The projection terms collapse each B-scan into 1D profiles—maximum and average along depth (axial) and along the lateral axis—and compare the reconstruction's profiles to ground truth with L1. This makes the network judge each slice by how it aggregates into en face views and by how vessel structure lines up across neighboring slices, which is how the method avoids bandi

What would settle it

Present the same trained model with volumes containing partial-slice blur, sheared, or non-contiguous corruption instead of whole contiguous B-scan dropouts. If VAMOS-OCTA's advantage over the vessel-weighted baseline disappears or its en face maps show the same banding and fragmentation, then the multi-axis supervision is tuned to the synthetic dropout model rather than to real handheld motion.

Watch

Extended reading notes

Core claim

The paper's central claim is that vessel-weighted intensity reconstruction alone over-smooths OCTA slices and fragments vessels in en face views, and that adding projection-consistency losses along two orthogonal axes fixes both defects. With axial projection supervision alone, en face maximum-intensity-projection accuracy jumps (SSIM from 0.763 to 0.888, NCC from 0.812 to 0.914, L1 roughly halved), but B-scan sharpness barely changes; adding lateral projection supervision is what removes horizontal banding and lifts perceptual B-scan metrics (LPIPS from 0.608 to 0.510, Sobel edge preservation from 0.313 to 0.427). The same training recipe, learned on synthetic contiguous slice dropouts, is

Load-bearing premise

The load-bearing premise is that dropping a center B-scan plus a random contiguous block of 1–6 neighboring B-scans, with block size drawn from a geometric distribution with p=0.4, faithfully reproduces the motion artifacts seen in real handheld OCTA; if real artifacts involve partial-slice blur or shearing rather than whole-slice dropouts, the claimed transfer to clinical volumes is unproven.

Editorial extensions

If this is right

En face MIP quality can be improved substantially by projection losses alone: with axial supervision, SSIM rises to 0.888 and NCC to 0.914 relative to the vessel-weighted baseline.
Lateral projection supervision is the component that removes horizontal banding and brings B-scan sharpness gains, so a complete volumetric fix needs both axes.
A 2.5D network with per-slice losses can produce 3D-consistent volumes, meaning explicit 3D architectures are not required for this inpainting task.
Performance degrades gracefully with corruption severity: mean intensity error stays low even when large contiguous blocks of slices are missing, where baseline methods' errors climb sharply.
The approach transfers to volumes with real-world motion corruptions, at least qualitatively, after training on synthetic contiguous slice dropouts.

Reading between the lines

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

A natural testable extension is to replace the L1 projection terms with perceptual or structural losses on the 1D profiles; if projection consistency is the operative constraint, stronger profile metrics could sharpen recovered vasculature further.
Because the corruption model only drops whole contiguous B-scans, the method's real-artifact performance would be clarified by stress-testing on partial-slice blur or shearing; that scenario is unaddressed in the paper.
The same loss recipe could be dropped into other slice-acquired 3D modalities, such as MRI or ultrasound volumes with out-of-plane motion, since it only requires collapsing volumes along two orthogonal axes; the paper hints at generalizability but does not demonstrate it.
A reader wanting clinical assurance might verify restored vessels with graph-connectivity or segmentation-continuity metrics, because the reported sharpness gains are perceptual and edge-based rather than anatomical.

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,

Quick take: this is a competent, incremental paper with one genuinely new piece — the VAMOS loss, which adds axial and lateral MIP/AIP projection terms to SOAD's vessel-weighted MSE for 2.5D OCTA B-scan inpainting. That specific recipe is not in the cited prior work, and the ablation (wMSE vs. wMSE+Axial vs. VAMOS) is the right way to show it. On LPIPS, Sobel edge preservation, Laplacian blur, and most MIP metrics, VAMOS clearly beats the baselines, and the qualitative figures look plausible. Credit where due: the experiments are clean, the ablations are properly staged, the corruption simulation is dynamic during training, and the code release is promised. The paper is honestly written about what it does.

Soft spots, in order of size. First, the synthetic corruption model only drops whole contiguous B-scans. Real handheld OCTA artifacts also include partial-slice blur, shearing, and intraslice distortion. The paper says training used 'synthetic and real-world corrupted volumes,' but Methods describe only synthetic corruption, and the real-world evidence is a single qualitative figure with no ground truth. So the abstract's 'consistently outperforms prior methods' is only supported for the dropout model. That is a real limitation, not a manufactured one.

Second, the paper's own Table 1 contradicts the 'without trade-offs' claim in Section 3.1. VAMOS is worse than SOAD on B-scan PSNR (26.122 vs. 27.084) and worse than axial-only on MIP PSNR (27.772 vs. 28.168). The authors emphasize perceptual metrics precisely because pixel-wise metrics favor blurring — that is a defensible choice — but they should not then claim no trade-offs. The PSNR dips are small but real, and ignoring them looks like cherry-picking.

Third, there is a partial circularity concern: the VAMOS loss directly minimizes MIP/AIP L1, and Table 1b reports MIP L1 and MIE as key gains. That does not invalidate the SSIM/NCC/PSNR results, which are not part of the loss, but the MIP L1 improvement is partly the training objective measuring itself.

Verdict: this paper deserves a serious referee. It is a solid incremental contribution to a narrow clinical problem, with sound methodology and honest ablations — the synthetic-to-real gap and the PSNR trade-offs are fixable with more careful claims and external validation. I would cite it if I worked on OCTA restoration, and I would send it to review, expecting revision rather than acceptance as-is.

Would I bring it to reading group? Maybe, if the group cares about medical-image inpainting loss design. Otherwise it is a 'skim the tables' paper.

Referee Report

4 major / 5 minor

Summary. The paper proposes VAMOS-OCTA, a 2.5D U-Net for inpainting motion-corrupted B-scans in handheld OCTA volumes. The method augments SOAD's vessel-weighted MSE loss with axial and lateral projection losses defined as per-B-scan 1D MIP/AIP L1 terms. The authors evaluate on a synthetic corruption model that drops contiguous blocks of B-scans and report improvements in B-scan perceptual metrics (LPIPS, Laplacian blur, Sobel edge preservation) and en face MIP metrics (L1, MIE, SSIM, NCC, PSNR) relative to standard MSE, SOAD, and an axial-only ablation. Qualitative results on real-world corrupted volumes are also shown.

Significance. If the central claims hold, the idea of using multi-axis projection supervision for slice-based volumetric inpainting is simple, practical, and potentially generalizable to other modalities. Strengths of the manuscript include the public code release, the dynamic synthetic corruption strategy during training, and the systematic ablation adding axial and lateral projection terms. The authors also correctly motivate the need to go beyond pixel-wise metrics by including perceptual and edge-aware B-scan evaluations. However, several load-bearing issues—especially a mismatch between the claimed en face supervision and the actual loss definition, circularity between the training loss and reported MIP metrics, and unsupported real-world generalization—currently prevent the paper from being accepted.

major comments (4)

[§2.3.2, Eq. (3) and Table 1b] The loss in Eq. (3) computes 1D projection profiles per B-scan: max over z gives a 1D profile in x, and max over x gives a 1D profile in z. The en face MIP evaluated in Table 1b is a 2D image formed by collapsing depth across the entire B-scan stack, i.e., a function of lateral position and B-scan index. The proposed loss never compares projections across B-scans, so the reported MIP improvements cannot be attributed to direct supervision of the en face MIP. If the authors intend to supervise en face projections, the loss should be computed on the 2D projection of the full volume or of the input stack; otherwise the mechanism described in the abstract and Section 2.3.2 should be revised.
[§2.3, Eq. (1)–(3) and Table 1b] There is a circularity concern. The VAMOS loss directly minimizes the L1 distance between predicted and ground-truth axial/lateral MIP and AIP profiles. Table 1b then reports MIP L1 and MIE as primary evidence of MIP improvement. Since MIP L1 is exactly the training objective and MIE is essentially the AIP L1 objective, these metrics are not independent measurements. The paper should either report projection metrics that are not aligned with the loss (e.g., vessel segmentation agreement, vessel skeleton connectivity, or a manual reading study) or clearly state that Table 1b is a sanity check rather than evidence of generalization.
[§3.1, Table 1 and Section 3.2] The claim that VAMOS-OCTA produces "consistent improvements ... without trade-offs" is contradicted by the paper's own numbers. In Table 1a, VAMOS-OCTA has lower B-scan PSNR (26.122 ± 0.709) than SOAD (27.084 ± 0.712) and standard MSE (27.036 ± 0.764). In Table 1b, VAMOS-OCTA has lower MIP PSNR (27.772 ± 0.865) than the axial-only ablation (28.168 ± 1.044). These PSNR losses may be acceptable if the authors value LPIPS and edge preservation more, but the text must acknowledge this trade-off and explain why PSNR is less relevant for this clinical task.
[§2.1.2 and Abstract] The abstract states that the model was "trained on both synthetic and real-world corrupted volumes," but the Methods section only describes synthetic corruption (contiguous whole-B-scan dropout). No real-world training data or procedure is described. Figure 4 shows qualitative real-world examples without ground truth, so the claim that VAMOS-OCTA "consistently outperforms prior methods" on real handheld motion artifacts is unsupported. Furthermore, the synthetic model only removes entire B-scans; real artifacts often include partial-slice blur, shearing, and non-contiguous missing regions. The authors should either add quantitative real-world validation (e.g., expert grading or a proxy metric on stable regions) or substantially temper the real-world claims in the abstract and conclusion.

minor comments (5)

[Table 1a] The table is titled "perceptual quality metrics" but includes a PSNR column, which is a pixel-wise fidelity metric. Please clarify the categorization or move PSNR elsewhere.
[§3.1] The statistical testing is not described in detail. Paired t-tests over 7 cross-validation folds are weak with n=7, and no multiple-comparison correction is mentioned. Please report effect sizes, confidence intervals, or per-fold results.
[§2.2] The architecture description is minimal: no number of layers, channels, normalization, or training hyperparameters (learning rate, epochs, optimizer) is given. The code link helps, but the paper should still include sufficient architectural detail for reproducibility.
[§3.1, Table 1b] MIE is never defined in the text. Please define Mean Intensity Error and state how it is computed.
[§2.3.2] The terms "axial" and "lateral" projections may confuse readers. Since each B-scan is 2D, "axial projection" collapses the depth dimension to produce a 1D profile along x, while "lateral projection" collapses x to produce a 1D profile along z. Please use more explicit names such as "depth-collapsed" and "lateral-collapsed" profiles.

Circularity Check

1 steps flagged · score 5.0 of 10

Reported en-face MIP L1 improvement is the training objective itself; the paper still has independent evidence in B-scan perceptual metrics and MIP SSIM/NCC.

fitted input called prediction [Section 2.3, Eq. (1) and Eq. (3); Section 3.1, Table 1b]
"L_VAMOS = L_wMSE + λ_proj ( L^(ax)_MIP + L^(lat)_MIP + L^(ax)_AIP + L^(lat)_AIP ) (1) ... L^(ax)_MIP = (1/W) Σ_x |max_z Yhat(x,z) − max_z Y(x,z)| (3) ... For en face MIP reconstruction, we use the pixel-wise accuracy metrics L1, MIE, SSIM, NCC, and PSNR to evaluate depth-integrated volumetric consistency."

The axial MIP loss in Eq. (3) is exactly the per-B-scan L1 distance between the predicted and ground-truth en face maximum-intensity projections; aggregating over the reconstructed B-scan stack makes it the Table 1b MIP L1 metric. VAMOS-OCTA is therefore trained to minimize the same quantity that is reported as its headline en-face improvement over SOAD, so that row is a restatement of the objective rather than an independent measurement. The MIE row is aligned with the AIP term as well. Independent support does exist (LPIPS/Sobel on B-scans, SSIM/NCC on MIPs), so the circularity is partial.

full rationale

The main load-bearing comparison is VAMOS-OCTA vs SOAD. The MIP L1 metric in Table 1b is not an independent probe of the method: it is the same functional as L^(ax)_MIP in the VAMOS loss (Eq. 3), and the improvement from 0.028 to 0.013 is expected by construction once that loss term is optimized. The same applies to MIE with respect to the AIP term. That said, the paper's central contribution is not solely MIP L1: the lateral projection loss, the B-scan LPIPS/Sobel results (0.510 vs 0.608 and 0.427 vs 0.313, both significant), and MIP SSIM/NCC improvements are not directly minimized and provide independent content. No load-bearing self-citation was found: SOAD [10] is an external method, and refs [24,25] by the same group are contextual. One support gap is flagged: the abstract states the model was 'trained on both synthetic and real-world corrupted volumes,' but Section 2.1.2 describes only synthetic corruption generation, and the only real-world evidence is qualitative (Figure 4, no ground truth). This is a verification gap, not circularity, but it weakens the claimed transfer to real handheld data. Overall, partial circularity on the MIP L1/MIE metrics, with independent evidence elsewhere: score 5.

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

The method rests on standard deep-learning assumptions plus a synthetic corruption model and a single-probe dataset. Hyperparameters lambda_proj=3, SOAD weights, and corruption parameters are chosen by hand. No new physical entities are introduced.

free parameters (4)

lambda_proj = 3
Weight balancing the projection losses against wMSE in Eq. 1; chosen by hand with no ablation reported.
wMSE constants alpha_w, gamma_w, c = 100, 1/3, 0.5
Taken from SOAD [10]; not fit here but they are hand-set hyperparameters that shape the main reconstruction loss.
corruption block distribution p and max length = p=0.4, max 6 slices
Geometric distribution for synthetic motion corruption, truncated to 6 B-scans. Chosen to mimic handheld motion patterns but not validated against real artifact statistics.
input stack size S = 9
Number of neighboring B-scans fed to the 2.5D U-Net; no ablation is reported for this hyperparameter.

assumptions (5)

domain assumption Synthetic contiguous block dropout (center + up to 6 neighbors) faithfully simulates real bulk-motion artifacts in handheld OCTA.
Section 2.1.2; the entire training and testing pipeline uses this corruption model, and real-world evidence is only qualitative.
domain assumption The registered, decorrelation-derived OCTA volumes used as ground truth are artifact-free enough to supervise and evaluate inpainting.
Section 2.1.1; SVD decorrelation and row-wise registration are themselves imperfect, so 'ground truth' is a processed estimate.
domain assumption Perceptual metrics (LPIPS, Laplacian blur, Sobel edge preservation) capture clinically meaningful reconstruction quality.
Section 3.1; used as primary evidence because pixel-wise metrics are said to favor blurring.
domain assumption A 2.5D U-Net can implicitly ignore corrupted input neighbors without an explicit validity mask.
Section 2.2; no mask is provided, so the network must learn validity discrimination from data.
domain assumption Seven volumes from one custom probe are representative of handheld OCTA across patients and sites.
Section 2.1.1; 5/1/1 split rotated over 7 folds, no external dataset is used.

how reviews work

0 comments

Cite this review

Pith. "Pith review of VAMOS-OCTA: Vessel-Aware Multi-Axis Orthogonal Supervision for Inpainting Motion-Corrupted OCT Angiography Volumes." pith.science (2026). https://pith.science/paper/Y3XKKOPZ

@misc{pith2026260200995,
  author       = {Pith},
  title        = {Pith review of: VAMOS-OCTA: Vessel-Aware Multi-Axis Orthogonal Supervision for Inpainting Motion-Corrupted OCT Angiography Volumes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y3XKKOPZ}},
  note         = {Machine review of arXiv:2602.00995}
}

read the original abstract

Handheld Optical Coherence Tomography Angiography (OCTA) enables noninvasive retinal imaging in uncooperative or pediatric subjects, but is highly susceptible to motion artifacts that severely degrade volumetric image quality. Sudden motion during 3D acquisition can lead to unsampled retinal regions across entire B-scans (cross-sectional slices), resulting in blank bands in en face projections. We propose VAMOS-OCTA, a deep learning framework for inpainting motion-corrupted B-scans using vessel-aware multi-axis supervision. We employ a 2.5D U-Net architecture that takes a stack of neighboring B-scans as input to reconstruct a corrupted center B-scan, guided by a novel Vessel-Aware Multi-Axis Orthogonal Supervision (VAMOS) loss. This loss combines vessel-weighted intensity reconstruction with axial and lateral projection consistency, encouraging vascular continuity in native B-scans and across orthogonal planes. Unlike prior work that focuses primarily on restoring the en face MIP, VAMOS-OCTA jointly enhances both cross-sectional B-scan sharpness and volumetric projection accuracy, even under severe motion corruptions. We trained our model on both synthetic and real-world corrupted volumes and evaluated its performance using both perceptual quality and pixel-wise accuracy metrics. VAMOS-OCTA consistently outperforms prior methods, producing reconstructions with sharp capillaries, restored vessel continuity, and clean en face projections. These results demonstrate that multi-axis supervision offers a powerful constraint for restoring motion-degraded 3D OCTA data. Our source code is available at https://github.com/MedICL-VU/VAMOS-OCTA.

Figures

Figures reproduced from arXiv: 2602.00995 by the authors.

**Figure 2.** Reconstructed B-scan for each method. Arrows highlight areas of interest: (IV) oversmooths vessels and (V) [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Depth-wise en face MIP reconstructions. (I) Ground Truth, (II) Corrupted, (III) Standard MSE, (IV) SOAD Weighted MSE, (V) wMSE + Axial, (VI) VAMOS-OCTA. Zoom panels highlight an area where (III), (IV), and (V) fail to remove artifacts, while VAMOS-OCTA (VI) restores a smooth and realistic projection with recovered vessels (arrows) [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: Mean Intensity Error (MIE) vs. corruption sever [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 2 linked inside Pith

[1]

Artifacts in optical coherence tomography angiography,

Anvari, P., Ashrafkhorasani, M., Habibi, A., and Falavarjani, K. G., “Artifacts in optical coherence tomography angiography,”Journal of Ophthalmic & Vision Research16(2), 271 (2021)

2021
[2]

Handheld optical coherence tomography angiography,

Yang, J., Liu, L., Campbell, J. P., Huang, D., and Liu, G., “Handheld optical coherence tomography angiography,” Biomedical optics express8(4), 2287–2300 (2017)

2017
[3]

Rapid volumetric oct image acquisition using compres- sive sampling,

Lebed, E., Mackenzie, P. J., Sarunic, M. V ., and Beg, M. F., “Rapid volumetric oct image acquisition using compres- sive sampling,” Optics express18(20), 21003–21012 (2010)

2010
[4]

Deep learning for motion artifact-suppressed octa image generation from both repeated and adjacent oct scans,

Lin, Z., Zhang, Q., Lan, G., Xu, J., Qin, J., An, L., and Huang, Y ., “Deep learning for motion artifact-suppressed octa image generation from both repeated and adjacent oct scans,” Mathematics12(3), 446 (2024)

2024
[5]

Deep-learning-based motion correction in optical coherence tomography angiography,

Li, A., Du, C., and Pan, Y ., “Deep-learning-based motion correction in optical coherence tomography angiography,” Journal of biophotonics14(12), e202100097 (2021)

2021
[6]

Make-a-volume: Leveraging latent diffusion mod- els for cross-modality 3d brain mri synthesis,

Zhu, L., Xue, Z., Jin, Z., Liu, X., He, J., Liu, Z., and Yu, L., “Make-a-volume: Leveraging latent diffusion mod- els for cross-modality 3d brain mri synthesis,” in [ International Conference on Medical Image Computing and Computer-Assisted Intervention], 592–601, Springer (2023)

2023
[7]

A two-stage framework for optical coherence tomography angiography image quality improvement,

Cao, J., Xu, Z., Xu, M., Ma, Y ., and Zhao, Y ., “A two-stage framework for optical coherence tomography angiography image quality improvement,” Frontiers in medicine10, 1061357 (2023)

2023
[8]

Signal attenuation-compensated projection-resolved oct angiography,

Wang, J., Hormel, T. T., Bailey, S. T., Hwang, T. S., Huang, D., and Jia, Y ., “Signal attenuation-compensated projection-resolved oct angiography,”Biomedical Optics Express14(5), 2040–2054 (2023)

Show all 25 references

[9]

Enhanced visualization of retinal microvascu- lature in optical coherence tomography angiography imaging via deep learning,

Kadomoto, S., Uji, A., Muraoka, Y ., Akagi, T., and Tsujikawa, A., “Enhanced visualization of retinal microvascu- lature in optical coherence tomography angiography imaging via deep learning,” Journal of Clinical Medicine9(5), 1322 (2020)

2020
[10]

Self-supervised denoising and bulk motion artifact removal of 3d optical coherence tomography angiography of awake brain,

Li, Z., Ren, J., Zou, Z., Garigapati, K., Du, C., Pan, Y ., and Ling, H., “Self-supervised denoising and bulk motion artifact removal of 3d optical coherence tomography angiography of awake brain,” in [ International Conference on Medical Image Computing and Computer-Assisted ...

2024
[11]

Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising,

Zhang, K., Zuo, W., Chen, Y ., Meng, D., and Zhang, L., “Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising,” IEEE transactions on image processing26(7), 3142–3155 (2017)

2017
[12]

Speckle2self: Learning self-supervised despeckling with attention mechanism for sar images,

Lin, H., Su, X., Zeng, Z., Xing, C., and Yin, J., “Speckle2self: Learning self-supervised despeckling with attention mechanism for sar images,” Remote Sensing17(23), 3840 (2025)

2025
[13]

Point-of-care widefield retinal octa mosaicking with handheld spec- trally encoded coherence tomography and reflectometry,

Watson, J. J., Xu, R., Hecht, R., and Tao, Y ., “Point-of-care widefield retinal octa mosaicking with handheld spec- trally encoded coherence tomography and reflectometry,” in [Optical Coherence Tomography and Coherence Domain Optical Methods in Biomedicine XXIX], PC1330507, S...

2025
[14]

Artifacts and artifact removal in optical coherence tomographic angiography,

Hormel, T. T., Huang, D., and Jia, Y ., “Artifacts and artifact removal in optical coherence tomographic angiography,” Quantitative Imaging in Medicine and Surgery11(3), 1120 (2021)

2021
[15]

Panretinal handheld oct angiography for pediatric retinal imaging,

Ni, S., Liang, G. B., Ng, R., Ostmo, S., Jia, Y ., Chiang, M. F., Huang, D., Skalet, A. H., Young, B. K., Campbell, J. P., et al., “Panretinal handheld oct angiography for pediatric retinal imaging,” Biomedical Optics Express15(5), 3412–3424 (2024)

2024
[16]

U-net: Convolutional networks for biomedical image segmentation,

Ronneberger, O., Fischer, P., and Brox, T., “U-net: Convolutional networks for biomedical image segmentation,” in [International Conference on Medical image computing and computer-assisted intervention], 234–241, Springer (2015)

2015
[17]

Spockmip: Segmentation of vessels in mras with enhanced continuity using maximum intensity projection as loss,

Radhakrishna, C., Chintalapati, K. V ., Kumar, S. C. H. R., Sutrave, R., Mattern, H., Speck, O., N ¨urnberger, A., and Chatterjee, S., “Spockmip: Segmentation of vessels in mras with enhanced continuity using maximum intensity projection as loss,” arXiv preprint arXiv:2407.086...

2024 arXiv
[18]

The unreasonable effectiveness of deep features as a perceptual metric,

Zhang, R., Isola, P., Efros, A. A., Shechtman, E., and Wang, O., “The unreasonable effectiveness of deep features as a perceptual metric,” in [Proceedings of the IEEE conference on computer vision and pattern recognition], 586–595 (2018)

2018
[19]

Image quality assessment: from error visibility to structural similarity,

Wang, Z., Bovik, A. C., Sheikh, H. R., and Simoncelli, E. P., “Image quality assessment: from error visibility to structural similarity,”IEEE transactions on image processing13(4), 600–612 (2004)

2004
[20]

A comprehensive survey of loss functions and metrics in deep learning,

Terven, J., Cordova-Esparza, D.-M., Romero-Gonz ´alez, J.-A., Ram ´ırez-Pedraza, A., and Ch ´avez-Urbiola, E., “A comprehensive survey of loss functions and metrics in deep learning,” Artificial Intelligence Review58(7), 195 (2025)

2025
[21]

Generating high-resolution 3d ct with 12-bit depth using a diffusion model with adjacent slice and intensity calibration network,

Jeong, J., Kim, K. D., Nam, Y ., Cho, K., Kang, J., Hong, G.-S., and Kim, N., “Generating high-resolution 3d ct with 12-bit depth using a diffusion model with adjacent slice and intensity calibration network,” in [ International Conference on Medical Image Computing and Comput...

2023
[22]

Imfine: 3d inpainting via geometry-guided multi-view refinement,

Shi, Z., Huo, D., Zhou, Y ., Min, Y ., Lu, J., and Zuo, X., “Imfine: 3d inpainting via geometry-guided multi-view refinement,” in [Proceedings of the Computer Vision and Pattern Recognition Conference], 26694–26703 (2025)

2025
[23]

The role of deep learning in medical image inpainting: A systematic review,

Santos, J. C., Tom ´as Pereira Alexandre, H., Seoane Santos, M., and Henriques Abreu, P., “The role of deep learning in medical image inpainting: A systematic review,”ACM Transactions on Computing for Healthcare6(3), 1–24 (2025)

2025
[24]

Self-supervised csf inpainting with synthetic atrophy for improved accuracy validation of cortical surface analyses,

Wang, J., Larson, K. E., and Oguz, I., “Self-supervised csf inpainting with synthetic atrophy for improved accuracy validation of cortical surface analyses,” arXiv preprint arXiv:2303.05777 (2023)

2023 arXiv
[25]

Robust multiple sclerosis lesion inpainting with edge prior,

Zhang, H., Bakshi, R., Bagnato, F., and Oguz, I., “Robust multiple sclerosis lesion inpainting with edge prior,” in [International Workshop on Machine Learning in Medical Imaging], 120–129, Springer (2020)

2020

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

Artifacts in optical coherence tomography angiography,

Anvari, P., Ashrafkhorasani, M., Habibi, A., and Falavarjani, K. G., “Artifacts in optical coherence tomography angiography,”Journal of Ophthalmic & Vision Research16(2), 271 (2021)

2021

[2] [2]

Handheld optical coherence tomography angiography,

Yang, J., Liu, L., Campbell, J. P., Huang, D., and Liu, G., “Handheld optical coherence tomography angiography,” Biomedical optics express8(4), 2287–2300 (2017)

2017

[3] [3]

Rapid volumetric oct image acquisition using compres- sive sampling,

Lebed, E., Mackenzie, P. J., Sarunic, M. V ., and Beg, M. F., “Rapid volumetric oct image acquisition using compres- sive sampling,” Optics express18(20), 21003–21012 (2010)

2010

[4] [4]

Deep learning for motion artifact-suppressed octa image generation from both repeated and adjacent oct scans,

Lin, Z., Zhang, Q., Lan, G., Xu, J., Qin, J., An, L., and Huang, Y ., “Deep learning for motion artifact-suppressed octa image generation from both repeated and adjacent oct scans,” Mathematics12(3), 446 (2024)

2024

[5] [5]

Deep-learning-based motion correction in optical coherence tomography angiography,

Li, A., Du, C., and Pan, Y ., “Deep-learning-based motion correction in optical coherence tomography angiography,” Journal of biophotonics14(12), e202100097 (2021)

2021

[6] [6]

Make-a-volume: Leveraging latent diffusion mod- els for cross-modality 3d brain mri synthesis,

Zhu, L., Xue, Z., Jin, Z., Liu, X., He, J., Liu, Z., and Yu, L., “Make-a-volume: Leveraging latent diffusion mod- els for cross-modality 3d brain mri synthesis,” in [ International Conference on Medical Image Computing and Computer-Assisted Intervention], 592–601, Springer (2023)

2023

[7] [7]

A two-stage framework for optical coherence tomography angiography image quality improvement,

Cao, J., Xu, Z., Xu, M., Ma, Y ., and Zhao, Y ., “A two-stage framework for optical coherence tomography angiography image quality improvement,” Frontiers in medicine10, 1061357 (2023)

2023

[8] [8]

Signal attenuation-compensated projection-resolved oct angiography,

Wang, J., Hormel, T. T., Bailey, S. T., Hwang, T. S., Huang, D., and Jia, Y ., “Signal attenuation-compensated projection-resolved oct angiography,”Biomedical Optics Express14(5), 2040–2054 (2023)