From Transthoracic to Transesophageal: Cross-Modality Generation using LoRA Diffusion

Anna Novikova; Daria Kulikova; Emmanuel Oladokun; Jurica \v{S}prem; Sarina Thomas; Vicente Grau; Yuxuan Ou

REVIEW 4 major objections 6 minor 25 references

From Transthoracic to Transesophageal: Cross-Modality Generation using LoRA Diffusion

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 A mask-conditioned diffusion model pretrained on transthoracic echo can be adapted to transesophageal echo with low-rank adapters and only 196 real TEE image-mask pairs, generating synthetic frames that improve right-heart segmentation…

desk verdict Useful practical adaptation of a TTE diffusion model to TEE with small LoRA adapters and mask remapping, but the abstract's 'MLP-only' claim conflicts with the methods and the train/test split may leak video identities. read the letter →

arxiv 2508.13077 v1 pith:EIM3PN6Z submitted 2025-08-18 eess.IV cs.AI

Emmanuel Oladokun , Yuxuan Ou , Anna Novikova , Daria Kulikova , Sarina Thomas , Jurica Šprem , Vicente Grau This is my paper · ORCID

classification eess.IVcs.AI

keywords low-rankadaptationdiffusionmodelstransesophagealechocardiographycross-modalitygenerationmaskconditioningsyntheticdataaugmentationcardiacsegmentationright-heartstructures

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

Deep diffusion models generate realistic images but normally need large training sets, and transesophageal echocardiography (TEE) has very little public data. The paper claims that a diffusion model pretrained on the common transthoracic view (TTE) can be moved to TEE using low-rank adapters of about $10^5$ to $5\times10^5$ parameters plus a mask-remapping layer, with only 196 real TEE image-mask pairs for training. The adapted model produces synthetic TEE frames that, when mixed with real frames, improve multiclass segmentation Dice, particularly for right-heart structures. The practical stake is a cheap route to controlled synthetic data in a modality where deep learning research is currently scarce.

What carries the argument

The machinery is a frozen, mask-conditioned Elucidated Diffusion Model (EDM) pretrained on TTE, plus two lightweight additions: LoRA adapters, which factor each weight update into low-rank products $BA$ and leave the base weights frozen, and MaskR$^2$, a remapping layer that converts a new dataset's semantic labels into the base model's one-hot conditioning channels using Identity, Reduce, and Repurpose operations. A targeted layer-group study attaches adapters independently to cross-attention, self-attention, convolution, linear, and other layers to show which carry the adaptation. This combination is what lets a model trained on left-heart TTE labels generate TEE images conditioned on masks that include right-heart structures.

What would settle it

Run the same LoRA and MaskR$^2$ adaptation protocol but initialize the frozen backbone with random weights instead of TTE-pretrained weights; if the random-backbone version still produces synthetic TEE frames that yield comparable segmentation Dice gains, then transfer from TTE is not what carries the result.

Watch

Extended reading notes

Core claim

The claim, on the paper's own terms, is that a frozen EDM diffusion backbone pretrained on TTE semantic-mask pairs can be adapted to TEE with LoRA adapters attached to selected layer groups and a MaskR$^2$ layer that maps TEE's mask labels (including right-heart structures) into the TTE model's conditioning channels. With 196 training TEE pairs and roughly $10^5$ to $5\times10^5$ trainable parameters, the adapted model generates anatomically faithful synthetic TEE images; the paper further claims that adapting the linear/MLP layers, with cross-attention adapters always present, is sufficient for high-fidelity synthesis. When the synthetic frames are mixed 1:1 with real TEE frames to train a segmentation network, global and class-weighted Dice improve over training on real frames alone, particularly for underrepresented right-heart structures. The paper also reports that image-quality scores such as FID, LPIPS, and SSIM correlate only weakly with this downstream segmentation benefit.

Load-bearing premise

The load-bearing premise is that the features learned from transthoracic echo transfer well enough to transesophageal echo that a few adapter layers trained on 196 TEE pairs capture the TEE-specific appearance (probe geometry, speckle texture, near-field artifacts) needed to produce realistic, useful synthetic images.

Editorial extensions

If this is right

With a frozen TTE-pretrained diffusion model and about $10^5$ to $5\times10^5$ trainable parameters, a research group can generate synthetic TEE frames after collecting only 196 annotated TEE image-mask pairs.
MaskR$^2$ lets the adapted model accept masks with new anatomical structures, such as right-heart chambers, by remapping them into the base model's existing conditioning channels rather than retraining the architecture.
Mixing synthetic TEE frames with real frames in a 1:1 ratio improves multiclass segmentation Dice, with the largest gains on the right-heart classes that are underrepresented in the real training data.
The targeted layer study implies that most adaptation capacity is concentrated in a few layer groups, so the adapters can be merged back into the base model after training and add no extra inference cost.
Because conditioning masks can be drawn from publicly available statistical shape models, the augmentation pipeline does not require a large private TEE dataset to reproduce.

Reading between the lines

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

The Evaluation section itself cautions that FID, LPIPS, and SSIM are built for natural scenes, so the reported image-quality scores should not be read as a complete measure of echo realism; the downstream segmentation gain is the more decision-relevant number.
The abstract's statement that 'adapting only MLP layers suffices' should be read alongside the Methods section's statement that cross-attention adapters are always trained; an ablation that removes cross-attention adaptation entirely would show where the adaptation capacity actually lives.
MaskR$^2$'s Identity/Reduce/Repurpose recipe is a general mechanism for moving any mask-conditioned generator to a dataset with a different label set, so the same pipeline could be tested for other ultrasound views or other imaging modalities whenever a suitable pretrained backbone exists.
The paper tests one training-set size (196 frames); an explicit curve showing how segmentation gain changes as the number of real TEE frames shrinks would map the data-efficiency limit of the approach.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Short version: the paper does something useful and mostly honest, but the abstract overstates one finding and the data split is under-described in a way that matters for the headline result.

The concrete contribution is a pipeline: take a TTE-trained mask-conditioned diffusion model (EDM/UNet), freeze it, attach LoRA adapters to selected layer groups, add a MaskR2 remapping layer to translate TEE masks into the base model's class channels, and fine-tune on 196 real TEE image-mask pairs. The authors then show that synthetic TEE images from this adapted model, mixed 1:1 with real frames, improve nnUNet multiclass Dice over real frames alone, with the largest gains on right-heart structures. That is a practical result for a clinically important but data-scarce modality. MaskR2 is simple and well-motivated, and the paper is right to evaluate with downstream segmentation rather than only FID/LPIPS, whose limits on ultrasound they acknowledge.

Now the soft spots. The abstract's claim that 'adapting only MLP layers suffices' runs head-on into the Methods sentence that cross-attention layers are always trained. If every adapter configuration includes cross-attention LoRA, the MLP-only result isn't actually MLP-only. This matters because the targeted adaptation analysis is one of the three listed contributions. The authors need to either correct the abstract or show a configuration that excludes cross-attention.

Second, the split. The 288 frames come from 71 videos, and the split is described only as balanced across views and phases. There is no statement that all frames from a video stay in one split. If training and test draw from the same videos, the segmentation gain could be inflated by memorized patient-specific speckle or anatomy. The paper also gives no error bars on the Dice numbers. This is fixable with a per-video grouped split and confidence intervals, but it's load-bearing for the main claim.

Minor: no code/data release (SSM masks are public, but the TEE data and model aren't), and the FID/LPIPS caveat is handled honestly.

Who's this for: people working on synthetic ultrasound generation, cross-modality adaptation, or label-efficient fine-tuning of diffusion models. It deserves a real referee, because the idea is solid and the results are plausible, but the current version needs the split clarified and the internal inconsistency resolved.

Recommendation: send to peer review, with a request for revision on those two points.

Referee Report

4 major / 6 minor

Summary. The paper proposes a pipeline that freezes an EDM diffusion model pretrained on CAMUS TTE mask-conditioned images and adapts it to TEE using LoRA adapters trained on only 196 real TEE image–mask pairs. A new MaskR2 layer remaps the TEE label set (LA, LV, RA, RV) into the base model's original TTE label space, so the conditioning architecture can be reused without retraining. The authors compare different LoRA adapter configurations on FID, LPIPS, and SSIM, and then train nnUNet segmentation models on real TEE frames mixed with synthetic images generated from either real masks or SSM-derived masks, reporting improvements in multiclass Dice, especially for right-heart structures. The abstract further claims that adapting only MLP layers suffices, while the methods state that cross-attention layers are always trained, an inconsistency that the paper does not resolve.

Significance. If the quantitative claims hold, this is a useful contribution to data-scarce echocardiography: it demonstrates parameter-efficient cross-modality adaptation with fewer than 200 real TEE pairs, proposes a simple MaskR2 label-remapping mechanism, and validates synthetic augmentation on a real downstream segmentation task rather than only on image-quality metrics. The use of external benchmarks (FID, LPIPS, downstream segmentation on real TEE) is a strength and avoids circularity. However, the version I reviewed does not establish data independence between training and test videos, and the internal contradiction about which layers are trained undermines the targeted-adaptation claim; both issues need to be addressed before the results can be considered reliable.

major comments (4)

[Methods, Data paragraph] The TEE set contains 288 image–mask pairs drawn from 71 videos, but the split into 196/40/52 images is described only as ensuring an even distribution across views and cardiac phases. There is no statement that all frames from a given video or patient are confined to a single split. If frames from the same video appear in both the training and test sets, the diffusion model and the downstream nnUNet can memorize video-specific speckle, probe geometry, and anatomy, inflating the reported FID/LPIPS and Dice gains. Please provide a video-exclusive split, ideally at the patient level, and report per-video grouped statistics with confidence intervals. Without this, the central augmentation claim is not supported.
[Abstract; Methods, LoRA strategy] The abstract states that 'adapting only MLP layers suffices for high-fidelity TEE synthesis,' but the Methods state that 'Cross-attention layers are always trained, as they control how the model integrates conditioning signals.' These statements are inconsistent: if cross-attention adapters are always trained, then no configuration actually tests MLP-only adaptation. Please either add an experimental configuration with only MLP LoRA adapters (and no cross-attention training) or revise the claim to say that among the optional adapter groups, adding MLP adapters yields the best trade-off. The exact parameter count of the minimal configuration should also be reported explicitly.
[Methods, Evaluation; Results & Discussion] The 'All-Weights' baseline is described as a model trained from scratch on the TEE dataset, yet the results use it to conclude that 'adapters are able to leverage the base model's prior knowledge from TTE data.' A from-scratch baseline cannot isolate the contribution of TTE pretraining; the appropriate control is a fully fine-tuned version of the TTE-pretrained backbone. If 'All-Weights' is actually a fully fine-tuned pretrained model, the text and the Figure 2 caption should say so explicitly. Please add this control or soften the attribution to TTE knowledge.
[Results & Discussion, Table 1 and Figure 2] The text repeatedly refers to 'Table 1' and 'Figure 2' for FID, LPIPS, SSIM, Dice, and parameter counts, but neither the table nor the figure appears in the manuscript text provided for review. Without the actual numbers, the central quantitative claims cannot be verified. Please ensure that the table and figure are included in the submission and are legible, or state the key numbers directly in the text.

minor comments (6)

[Abstract; Results & Discussion] The abstract says 'adapters as small as 10^5 parameters' while the Results say 'around 510,000 parameters'; please use consistent notation and specify the exact minimal parameter count for the best configuration.
[Results & Discussion, image-quality metrics] The FID, LPIPS, and SSIM values and the claimed weak correlation with segmentation impact are reported without confidence intervals; with only 52 test images, FID is high-variance, so consider reporting bootstrap intervals or multiple sampling draws, or explicitly label the correlation analysis as exploratory.
[Methods, Diffusion formulation] Equation (2) uses an unusual underbrace notation around the network output; please rewrite the loss to be fully explicit so that the reader can follow the EDM preconditioning without ambiguity.
[References] Reference [4] for LoRA lacks a venue and year, and reference [6] is cited only as an arXiv preprint; please update these entries to their published versions where available.
[Figure 2 caption] The term 'All Weights' is used for a model said to be trained from scratch, but the main text refers to it as a baseline for LoRA adaptation; using distinct names such as 'FromScratch' and 'FullFT' would reduce ambiguity.
[Data availability] The internal TEE dataset is not described as available, which limits reproducibility; adding a data availability statement or a plan for releasing de-identified data would be helpful.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the LoRA adaptation and downstream segmentation gains are evaluated against external benchmarks and held-out real TEE frames; the only self-citation (SSM mask pipeline) is a non-load-bearing data source.

full rationale

The paper's derivation chain is self-contained rather than circular. The LoRA adapters are trained on real TEE image–mask pairs, and the central empirical claims are evaluated with external metrics (FID, LPIPS, SSIM) on a held-out test set and with downstream nnUNet segmentation on held-out real TEE frames against a real-only baseline. The segmentation Dice gains are not constructed from the generative model's own outputs; they are measured on real images. MaskR2 is a label-remapping layer, not a quantity derived from the target results. The only author-overlapping citation is the SSM mask pipeline [11], used to supply out-of-distribution masks for synthetic augmentation. That is a source of conditioning inputs, not a derived prediction, and the paper also mixes synthetic images generated from real masks in a 1:1 ratio, so the central claim does not reduce to the cited pipeline. The manuscript's lack of a video-exclusive data split is a potential data-leakage or correctness risk, but it is not a circularity: it does not make the reported outcomes equivalent to the method's inputs by construction. The contradiction between the abstract's 'adapting only MLP layers suffices' and the methods' 'Cross-attention layers are always trained' is an internal-reporting inconsistency, not a circular step. Overall, no prediction or first-principles result is definitionally forced by the inputs.

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

The method relies on standard diffusion and LoRA machinery plus the authors' MaskR2 remapping. The key free parameters are training hyperparameters chosen by search; the central scientific assumption is cross-modality transferability of a frozen backbone.

free parameters (5)

LoRA rank and alpha = r = alpha = 16
Chosen via hyperparameter search; controls capacity of adapters.
Learning rate = 1e-3 with cosine decay
Selected by hyperparameter search for all adapted models.
Training steps = 100,000
Fixed training budget for all adapted models.
Batch size = 4
Fixed batch size for training.
Augmentation mixing ratio = 1:1 real to synthetic
Chosen for downstream segmentation experiments; not ablated.

assumptions (4)

standard math EDM diffusion formulation and LoRA low-rank decomposition work as described in prior literature.
The paper adopts these frameworks without modification.
domain assumption TTE and TEE share enough anatomical structure and ultrasound appearance for transfer learning.
The entire adaptation strategy depends on this assumption.
domain assumption Merging right-heart structures into a single super-class in MaskR2 preserves enough conditioning information for segmentation.
The paper uses this mapping for all synthetic images, but does not ablate alternative mappings.
ad hoc to paper MaskR2's three operations (Identity, Reduce, Repurpose) can map any new label set into the base model's label space without retraining the conditioning architecture.
This is proposed by the authors and not proven, only demonstrated on the described TEE-to-CAMUS mapping.

how reviews work

0 comments

Cite this review

Pith. "Pith review of From Transthoracic to Transesophageal: Cross-Modality Generation using LoRA Diffusion." pith.science (2026). https://pith.science/paper/EIM3PN6Z

@misc{pith2026250813077,
  author       = {Pith},
  title        = {Pith review of: From Transthoracic to Transesophageal: Cross-Modality Generation using LoRA Diffusion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EIM3PN6Z}},
  note         = {Machine review of arXiv:2508.13077}
}

abstract

Deep diffusion models excel at realistic image synthesis but demand large training sets-an obstacle in data-scarce domains like transesophageal echocardiography (TEE). While synthetic augmentation has boosted performance in transthoracic echo (TTE), TEE remains critically underrepresented, limiting the reach of deep learning in this high-impact modality. We address this gap by adapting a TTE-trained, mask-conditioned diffusion backbone to TEE with only a limited number of new cases and adapters as small as $10^5$ parameters. Our pipeline combines Low-Rank Adaptation with MaskR$^2$, a lightweight remapping layer that aligns novel mask formats with the pretrained model's conditioning channels. This design lets users adapt models to new datasets with a different set of anatomical structures to the base model's original set. Through a targeted adaptation strategy, we find that adapting only MLP layers suffices for high-fidelity TEE synthesis. Finally, mixing less than 200 real TEE frames with our synthetic echoes improves the dice score on a multiclass segmentation task, particularly boosting performance on underrepresented right-heart structures. Our results demonstrate that (1) semantically controlled TEE images can be generated with low overhead, (2) MaskR$^2$ effectively transforms unseen mask formats into compatible formats without damaging downstream task performance, and (3) our method generates images that are effective for improving performance on a downstream task of multiclass segmentation.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 14 canonical work pages

[1]

Science Robotics 3(January), 2672–2680 (2014)

Goodfellow, I.J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y.: Generative Adversarial Networks. Science Robotics 3(January), 2672–2680 (2014)

work page 2014
[2]

Journal of the American Society of Echocardiography 35, 1–76 (2022)

Hahn, R.T., Saric, M., Faletra, F.F., Garg, R., Gillam, L.D., Horton, K., Khalique, O.K., Little, S.H., Mackensen, G.B., Oh, J., Quader, N., Safi, L., Scalia, G.M., Lang, R.M.: Recommended Standards for the Performance of Transesophageal EchocardiographicScreeningforStructuralHeartIntervention:FromtheAmerican Society of Echocardiography. Journal of the Am...

work page doi:10.1016/j.echo.2021.07.006 2022
[3]

Advances in Neural Information Processing Systems2017-December, 6627–6638 (2017)

Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Hochreiter, S.: GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium. Advances in Neural Information Processing Systems2017-December, 6627–6638 (2017). https://doi.org/10.18034/ajase.v8i1.9

doi:10.18034/ajase.v8i1.9 2017
[4]

Hu, E., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., Chen, W.: LORA: LOW-RANK ADAPTATION OF LARGE LAN-GUAGE MODELS

work page
[5]

Nature Methods 18, 203–211 (2021)

Isensee, F., Jaeger, P.F., Kohl, S.A., Petersen, J., Maier-Hein, K.H.: nnU-Net: a self-configuring method for deep learning-based biomedi- cal image segmentation. Nature Methods 18, 203–211 (2021). https: //doi.org/10.1038/S41592-020-01008-Z;SUBJMETA=114,1564,308,575,631, 692;KWRD=IMAGE+PROCESSING,TRANSLATIONAL+RESEARCH

work page doi:10.1038/s41592-020-01008-z;subjmeta 2021
[6]

Karras, T., Aittala, M., Aila, T., Laine, S.: Elucidating the Design Space of Diffusion-Based Generative Models (2022), http://arxiv.org/abs/2206.00364

arXiv 2022
[7]

2nd International Conference on Learning Representations, ICLR 2014 - Conference Track Proceed- ings (2013)

Kingma, D.P., Welling, M.: Auto-Encoding Variational Bayes. 2nd International Conference on Learning Representations, ICLR 2014 - Conference Track Proceed- ings (2013). https://doi.org/10.61603/ceas.v2i1.33

doi:10.61603/ceas.v2i1.33 2013
[8]

IEEE transactions on medical imag- ing 38(9), 2198–2210 (9 2019)

Leclerc, S., Smistad, E., Pedrosa, J., Ostvik, A., Cervenansky, F., Espinosa, F., Espeland, T., Berg, E.A.R., Jodoin, P.M., Grenier, T., Lartizien, C., Dhooge, J., Lovstakken, L., Bernard, O.: Deep Learning for Segmentation Using an Open Large-Scale Dataset in 2D Echocardiography. IEEE transactions on medical imag- ing 38(9), 2198–2210 (9 2019). https://d...

arXiv 2019

Show all 25 references

[9]

Magyar, B., Tokodi, M., Soos, A., Tolvaj, M., Lakatos, B.K., Fábián, A., Surkova, E., Merkely, B., Kovács, A., Horváth, A.: RVENet: A Large Echocardiographic Dataset for the Deep Learning-Based Assessment of Right Ventricular Function
[10]

Nguyen, V.P., Nhan, T., Ha, L., Pham, H.H., Long, Q.,†1, T.: Training-Free Con- dition Video Diffusion Models for single frame Spatial-Semantic Echocardiogram Synthesis (8 2024), https://arxiv.org/abs/2408.03035v2

2024 arXiv
[11]

Lecture Notes in Computer Sci- ence (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 14379 LNCS , 43–52 (2024)

Oladokun, E., Abdulkareem, M., Šprem, J., Grau, V.: Transesophageal Echocar- diography Generation using Anatomical Models. Lecture Notes in Computer Sci- ence (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 14379 LNCS , 43–52 ...

2024
[12]

In: NeurIPS ML4H Workshop: Vancouver, BC, Canada

Ouyang, D., He, B., Ghorbani, A., Lungren, M.P., Ashley, E.A., Liang, D.H., Zou, J.Y.: Echonet-dynamic: a large new cardiac motion video data resource for medical machine learning. In: NeurIPS ML4H Workshop: Vancouver, BC, Canada. vol. 5 (2019)

2019
[13]

Park, T., Efros, A.A., Zhang, R., Zhu, J.Y.: Contrastive Learning for Un- paired Image-to-Image Translation. Lecture Notes in Computer Science (in- cluding subseries Lecture Notes in Artificial Intelligence and Lecture Notes Title Suppressed Due to Excessive Length 11 in Bioin...

2020
[14]

Phil Wang: GitHub - lucidrains/imagen-pytorch: Implementation of Ima- gen, Google’s Text-to-Image Neural Network, in Pytorch, https://github.com/ lucidrains/imagen-pytorch
[15]

British Journal of General Practice 69(684), 358–359 (2019)

Potter, A., Pearce, K., Hilmy, N.: The benefits of echocardiography in primary care. British Journal of General Practice 69(684), 358–359 (2019). https://doi. org/10.3399/BJGP19X704513

2019 doi
[16]

https://doi.org/10.1007/978-3-031-43999-5{_}14

Reynaud, H., Qiao, M., Dombrowski, M., Day, T., Razavi, R., Gomez, A., Leeson, P., Kainz, B.: Feature-Conditioned Cascaded Video Diffusion Models for Precise EchocardiogramSynthesis.LectureNotes inComputerScience(includingsubseries Lecture Notes in Artificial Intelligence and ...

2023 doi
[17]

PLOS Computational Biology 17(4), e1008851 (4 2021)

Rodero, C., Strocchi, M., Marciniak, M., Longobardi, S., Whitaker, J., O’Neill, M.D., Gillette, K., Augustin, C., Plank, G., Vigmond, E.J., Lamata, P., Niederer, S.A.: Linking statistical shape models and simulated function in the healthy adult human heart. PLOS Computational ...

2021 doi
[18]

Ronneberger, O., Fischer, P., Brox, T.: U-Net: Convolutional Networks for Biomed- ical Image Segmentation. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)9351, 234–241 (2015), https://arxiv.or...

2015 arXiv
[19]

Sohl-Dickstein, J., Weiss, E.A., Maheswaranathan, N., Ganguli, S., Edu, S.: Deep Unsupervised Learning using Nonequilibrium Thermodynamics
[20]

Stojanovski, D., Gomez, A.: Echo from noise: synthetic ultrasound image genera- tion using diffusion models for realimage segmentation https://zenodo.org/record/ 7921055#.ZGYS
[21]

IEEE Transactions on Image Process- ing 13(4), 600–612 (2004)

Wang, Z., Bovik, A.C., Sheikh, H.R., Simoncelli, E.P.: Image quality assessment: From error visibility to structural similarity. IEEE Transactions on Image Process- ing 13(4), 600–612 (2004). https://doi.org/10.1109/TIP.2003.819861

2004
[22]

Lecture Notes in Computer Science (including subseries Lec- ture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)13438 LNCS, 35–45 (2022)

Wolleb, J., Bieder, F., Sandkühler, R., Cattin, P.C.: Diffusion Models for Medical Anomaly Detection. Lecture Notes in Computer Science (including subseries Lec- ture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)13438 LNCS, 35–45 (2022). https://doi.org...

2022 doi
[23]

Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition pp

Zhang, R., Isola, P., Efros, A.A., Shechtman, E., Wang, O.: The Unreasonable Effectiveness of Deep Features as a Perceptual Metric. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition pp. 586–595 (2018). https://doi.org/10.1109/CVPR.2...

2018
[24]

Medical Image Analysis (2024), www.elsevier.com/ locate/media

Zhou, Y., Chen, T., Hou, J., Xie, H., Dvornek, N.C., Zhou, S.K., Wilson, D.L., Duncan, J.S., Liu, C., Zhou, B.: Cascaded Multi-path Shortcut Diffusion Model for Medical Image Translation. Medical Image Analysis (2024), www.elsevier.com/ locate/media

2024
[25]

Proceedings of the IEEE In- ternational Conference on Computer Vision 2017-October, 2242–2251 (2017)

Zhu, J.Y., Park, T., Isola, P., Efros, A.A.: Unpaired Image-to-Image Transla- tion using Cycle-Consistent Adversarial Networks. Proceedings of the IEEE In- ternational Conference on Computer Vision 2017-October, 2242–2251 (2017). https://doi.org/10.1109/ICCV.2017.244

2017 doi

Pith tools

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

[1] [1]

Science Robotics 3(January), 2672–2680 (2014)

Goodfellow, I.J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y.: Generative Adversarial Networks. Science Robotics 3(January), 2672–2680 (2014)

work page 2014

[2] [2]

Journal of the American Society of Echocardiography 35, 1–76 (2022)

Hahn, R.T., Saric, M., Faletra, F.F., Garg, R., Gillam, L.D., Horton, K., Khalique, O.K., Little, S.H., Mackensen, G.B., Oh, J., Quader, N., Safi, L., Scalia, G.M., Lang, R.M.: Recommended Standards for the Performance of Transesophageal EchocardiographicScreeningforStructuralHeartIntervention:FromtheAmerican Society of Echocardiography. Journal of the Am...

work page doi:10.1016/j.echo.2021.07.006 2022

[3] [3]

Advances in Neural Information Processing Systems2017-December, 6627–6638 (2017)

Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Hochreiter, S.: GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium. Advances in Neural Information Processing Systems2017-December, 6627–6638 (2017). https://doi.org/10.18034/ajase.v8i1.9

doi:10.18034/ajase.v8i1.9 2017

[4] [4]

Hu, E., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., Chen, W.: LORA: LOW-RANK ADAPTATION OF LARGE LAN-GUAGE MODELS

work page

[5] [5]

Nature Methods 18, 203–211 (2021)

Isensee, F., Jaeger, P.F., Kohl, S.A., Petersen, J., Maier-Hein, K.H.: nnU-Net: a self-configuring method for deep learning-based biomedi- cal image segmentation. Nature Methods 18, 203–211 (2021). https: //doi.org/10.1038/S41592-020-01008-Z;SUBJMETA=114,1564,308,575,631, 692;KWRD=IMAGE+PROCESSING,TRANSLATIONAL+RESEARCH

work page doi:10.1038/s41592-020-01008-z;subjmeta 2021

[6] [6]

Karras, T., Aittala, M., Aila, T., Laine, S.: Elucidating the Design Space of Diffusion-Based Generative Models (2022), http://arxiv.org/abs/2206.00364

arXiv 2022

[7] [7]

2nd International Conference on Learning Representations, ICLR 2014 - Conference Track Proceed- ings (2013)

Kingma, D.P., Welling, M.: Auto-Encoding Variational Bayes. 2nd International Conference on Learning Representations, ICLR 2014 - Conference Track Proceed- ings (2013). https://doi.org/10.61603/ceas.v2i1.33

doi:10.61603/ceas.v2i1.33 2013

[8] [8]

IEEE transactions on medical imag- ing 38(9), 2198–2210 (9 2019)

Leclerc, S., Smistad, E., Pedrosa, J., Ostvik, A., Cervenansky, F., Espinosa, F., Espeland, T., Berg, E.A.R., Jodoin, P.M., Grenier, T., Lartizien, C., Dhooge, J., Lovstakken, L., Bernard, O.: Deep Learning for Segmentation Using an Open Large-Scale Dataset in 2D Echocardiography. IEEE transactions on medical imag- ing 38(9), 2198–2210 (9 2019). https://d...

arXiv 2019