Pith. sign in

REVIEW 3 major objections 4 minor 49 references

J-RAS: Mutual Adaptation for Medical Image Segmentation via Contrastive Retrieval-Augmented Joint Optimization

T0 review · 3 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read Jointly optimizing a retrieval model with a segmentation model improves medical image segmentation by making the retriever learn which image-mask pairs are most useful for delineating anatomy.

desk verdict The contribution—making the retriever trainable through the segmentation loss—is real and the gains are consistent, but the paper never runs the frozen-retriever control that would prove the joint update is what helps. read the letter →

arxiv 2510.09953 v3 pith:5KRCUXWV submitted 2025-10-11 cs.CV

Salma J. Ahmed , Emad A. Mohammed , Azam Asilian Bidgoli This is my paper · ORCID
classification cs.CV
keywords ImageSegmentationRetrievalJointTrainingRetrieval-AugmentedFeatureFusionMedicalContrastiveLearningCardiacMRI
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

J-RAS jointly trains a retrieval model and a segmentation model so that retrieved image-mask pairs guide the segmentation of new images. The retrieval network is first contrastively pretrained and then updated through the segmentation loss, learning to fetch cases that are anatomically relevant rather than merely visually similar. The paper shows this mutual adaptation improves Dice scores and reduces contour errors across four backbones (U-Net, TransUNet, SAM, SegFormer) and across cardiac MRI datasets, including a cross-dataset setting where guides come from one dataset and segmentation is done on another. If correct, this offers a path to better segmentation with less reliance on large labeled datasets.

What carries the argument

The central mechanism is a joint optimization loop: a radiology-pretrained DINOv2 vision transformer serves as the retrieval encoder, producing 128-dimensional embeddings for query and knowledge-base slices. Top-k retrieval uses cosine similarity; retrieved images and masks are weighted by a temperature-scaled softmax and fused into a composite image-mask pair, concatenated with the query into a 7-channel tensor, and projected via an adapter into the segmentation model. The same Dice-plus-cross-entropy loss then backpropagates through the similarity and weighting computations into the query branch, updating the retrieval encoder. Knowledge-base embeddings are recomputed each epoch.

What would settle it

An ablation study that keeps the retrieval encoder frozen after contrastive pretraining while using the identical top-k fusion and input representation: if the frozen-retrieval version matches or exceeds J-RAS's Dice and HD, the joint update is not the cause of the improvement.

Watch

Extended reading notes

Core claim

The central claim is that segmentation accuracy improves when the retrieval model is not fixed but co-optimized with the segmentation model. In J-RAS, a query image is embedded, top-k similar slices are retrieved from a knowledge base, and their images and ground-truth masks are merged into a composite guide that is concatenated with the query and fed to the segmentation network. The segmentation loss updates both the segmentation weights and the query branch of the retrieval encoder, so the encoder gradually retrieves examples that are more useful for delineating boundaries. On the ACDC test set, SegFormer's mean Dice rises from 0.8708±0.042 to 0.9115±0.031 and Hausdorff distance falls from

Load-bearing premise

The gains rest on the assumption that updating the retrieval model through the segmentation loss is what drives the improvement, rather than simply providing ground-truth masks to the segmentation network — an assumption not isolated by ablation in the paper.

Editorial extensions

If this is right

  • Segmentation models can tap into a growing knowledge base of annotated cases, improving boundary delineation without needing more labeled training data.
  • The retrieval model becomes a task-aware encoder, potentially useful for other downstream tasks like anomaly detection or case retrieval for clinical decision support.
  • Cross-dataset retrieval shows guidance can transfer across domains, suggesting applicability to low-resource settings where annotated data comes from a different scanner or protocol.
  • Consistent gains across CNN, hybrid, and transformer backbones imply the method is model-agnostic and can be layered onto existing segmentation pipelines.
  • Even noisy or corrupted guides outperform the no-guide baseline, indicating resilience to imperfect retrieval.

Reading between the lines

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

  • The paper does not ablate the joint update: a natural test is to freeze the retrieval encoder after contrastive pretraining and run the same top-k fusion. If Dice gains persist, the improvement may come from the extra ground-truth masks rather than mutual adaptation.
  • Because only the query branch is updated (gallery embeddings are detached per epoch), the retrieval model may develop a query-specific bias; a full end-to-end update of the knowledge base could improve consistency but is currently limited by computation.
  • The dynamic Top-K strategy underperformed fixed K=1 in their experiments; a learned or confidence-based selection rule may be a promising extension beyond the fixed threshold used here.
  • The cross-dataset results suggest the method could be extended to multiorgan or multimodal segmentation, where guides from one modality support segmentation in another.
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.

Referee Report

3 major / 4 minor

Summary. The paper proposes J-RAS, a retrieval-augmented segmentation method that couples a segmentation model (U-Net, TransUNet, SAM, SegFormer) with a DINOv2/RAD-DINO retrieval encoder. After an independent contrastive fine-tuning stage, the method retrieves the top-k most similar image–mask pairs, fuses them via temperature-scaled softmax weights, and concatenates the query image, fused guide image, and fused mask into a 7-channel input to the segmentation model. The segmentation loss is then backpropagated into both the segmentation model and the retrieval encoder's query branch, with knowledge-base embeddings recomputed each epoch. The paper reports consistent Dice/HD improvements over backbone-only baselines on ACDC and M&Ms, and includes ablations on fusion strategy, dynamic vs. fixed top-k, noisy guides, and cross-dataset retrieval.

Significance. If the mutual-adaptation loop works as stated, J-RAS would be a practically useful, model-agnostic way to inject retrieved anatomical context into medical image segmentation, especially for small datasets. The paper has genuine strengths: it evaluates four diverse backbones, includes a cross-dataset retrieval experiment, tests robustness to noisy guides, and provides qualitative evidence that retrieval quality changes after joint training. However, the central causal claim is not isolated by any experiment: no frozen-retriever baseline is reported under identical fusion and input conditions, and the gradient mechanism through the discrete top-k selection is not specified. The contributions are therefore plausible but currently under-supported.

major comments (3)
  1. [Section 3.2, Algorithm 1, Table 2] The central claim—that jointly optimizing the retrieval model R_theta with the segmentation loss improves segmentation—is not isolated. Every J-RAS result in Table 2 changes two factors relative to the backbone-only baseline: the segmentation input now contains a fused retrieved image-mask pair, and R_theta has been updated by the segmentation loss. There is no control that freezes R_theta after the independent contrastive stage while using the identical 7-channel early fusion, same softmax weights (Eq. 3), same top-k=2, and same training budget. The ablations in Section 6 vary fusion, k, noise, and dataset, but they all keep the joint gradient active, so they cannot separate the contribution of retrieval adaptation from the contribution of simply providing auxiliary ground-truth masks. Please add a frozen-retriever baseline; without it, the mutual-adaptation claim is untested.
  2. [Section 3.2, Algorithm 1] The gradient path for the joint update is underspecified. Algorithm 1 line 7 calls Retrieve(q, D'), a discrete top-k operation. The text states that the segmentation loss 'back-propagates through the similarity and weighting computations into the query embedding,' but no differentiable surrogate or straight-through estimator for the discrete selection is described. If the gradient flows only through the softmax weights w_i for already-selected indices, the dependence of the selected set on R_theta is ignored; if a relaxation is used, it should be stated explicitly. Since this gradient path is what makes the method 'joint,' the ambiguity is load-bearing for the method's correctness.
  3. [Abstract vs. Section 4.1] The arXiv abstract claims evaluation on 'four public benchmarks spanning different imaging modalities, including ACDC and M&Ms (MRI), Breast Cancer Ultrasound, and lung and infection CT,' but Section 4.1 and all experiments use only ACDC and M&Ms, both cardiac MRI. The paper body's own abstract says 'two benchmark datasets.' The generalizability claim is therefore overstated; either add the missing datasets or correct the abstract to match the actual experiments.
minor comments (4)
  1. [Table 2, SegFormer baseline row] The class-wise Dice scores (RV 0.8589, MYO 0.8484, LV 0.8484) do not average to the reported mean of 0.8708; the mean of the listed entries is 0.8519. This looks like a typo in one of the class entries and should be corrected, since readers use this table to verify the headline improvement.
  2. [Section 5.2, Figure 9] The text says that after J-RAS the model 'correctly identified slice 1 from the same patient as the closest match,' but Section 3.2 explicitly excludes slices from the same patient, and the figure labels show retrieved slices from different patients (e.g., P.113, P.147, P.120). The wording appears to be an error and should be fixed.
  3. [Section 5.1.2] No statistical significance tests are reported, and it is unclear whether results come from one run or multiple seeds. Given that some improvements (e.g., U-Net Dice 0.8588±0.040 to 0.9034±0.033) are small relative to the reported variance, paired significance tests across the 100 testing patients would strengthen the central claim.
  4. [Algorithm 1] Line 8 uses w_1 and w_2 without defining them in the algorithm; the definition in Eq. (3) should be referenced or repeated so the pseudocode is self-contained.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the paper reports an empirical joint-training loop, and its claims do not reduce to fitted constants or self-citations.

full rationale

J-RAS is an empirical optimization method, not a derivation from first principles. Algorithm 1 defines an iterative loop in which the segmentation loss is used to update both the retrieval and segmentation models (line 12: 'Update R_theta, S_phi'). The reported Dice/HD gains are measured outcomes of this loop, not quantities that are equal to the inputs by construction. Equations (3) and (4) define the fused guide as a similarity-weighted average; this is a design choice rather than a fitted parameter being renamed as a prediction. The paper's claim that the retrieval model 'learns segmentation-relevant features' is supported by segmentation improvements and qualitative retrieval examples, but this validation is empirical rather than circular: there is no equation in which the target result is assumed as an input. The absence of a frozen-retriever ablation is a real experimental confound — the gains may be due to the addition of ground-truth masks rather than to the retrieval-model update — but that is a causal-attribution problem, not a reduction of a prediction to its inputs. The paper itself acknowledges a related limitation in Section 7: 'Computational constraints prevented backpropagation through the entire retrieval knowledge base, restricting updates to the query side.' This narrows the claimed mutual adaptation but does not make the argument circular. There are no load-bearing self-citations, no 'uniqueness theorems' imported from the authors, and no ansatz smuggled in via citation. The external comparison with the retrieval-augmented few-shot method [43] and the per-backbone baselines give the evaluation independent empirical content. Accordingly, the circularity score is 0.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities or latent variables. The main extraneous assumptions are empirical design choices: the temperature parameters, the linear fusion of guides, and the untested premise that partial gradient flow through the query branch is sufficient to produce segmentation-aware retrieval.

free parameters (6)
  • Top-K (number of retrieved guides) = 2
    Selected based on Figure 2 showing performance saturates around K=2–6; fixed to 2 for the main experiments. This is a validation-set-driven choice and not a principled optimum.
  • Softmax temperature τ for fusion weights = not reported
    Used in Equation 3 to compute guide weights; no value is given anywhere in the paper.
  • NT-Xent temperature = not reported
    Used in contrastive loss (Eq. 1); value not reported.
  • Retrieval MLP projection dimensions = 768 → 256 → 128
    Architecture choice stated in Figure 1; no ablation.
  • Adapter for 7-to-3 channel projection = 1x1 convolution (implied)
    Described as 'lightweight adapter' in Section 3.2 and '1×1 convolution' in Section 6.1; no architectural detail or ablations.
  • Dynamic top-k bounds (k_min, k_max, threshold θ) = k_min=1, k_max=10, θ not reported
    Only used in the dynamic top-k ablation; not used in final method.
assumptions (5)
  • domain assumption DINOv2/RAD-DINO pretrained features are suitable for cardiac MRI retrieval.
    The retrieval encoder is initialized with RAD-DINO and fine-tuned; the paper provides no evidence that this initialization is optimal or that the domain shift to cardiac MRI is handled.
  • domain assumption Weighted averaging of retrieved images/masks preserves useful anatomical cues.
    Eq. 4 computes a linear combination of pixel images and masks; this presupposes that linear fusion of structures is meaningful for segmentation guidance.
  • ad hoc to paper Backpropagating the segmentation loss through the retrieval query branch steers retrieval toward more useful guides.
    This is the central mechanism of J-RAS, but no ablation isolates it from the general effect of adding retrieved ground-truth masks.
  • domain assumption Detaching knowledge base embeddings per epoch is a sufficient approximation to full end-to-end training.
    The paper states this is due to computational constraints; it limits the 'mutual adaptation' and is not benchmarked against full backpropagation.
  • domain assumption Ground-truth masks of other patients can be used as input at test time.
    The method requires annotated guides at inference; this is typical of retrieval-augmented methods but assumes such annotated examples are always available.

how reviews work

0 comments
Cite this review

Pith. "Pith review of J-RAS: Mutual Adaptation for Medical Image Segmentation via Contrastive Retrieval-Augmented Joint Optimization." pith.science (2026). https://pith.science/paper/5KRCUXWV

@misc{pith2026251009953,
  author       = {Pith},
  title        = {Pith review of: J-RAS: Mutual Adaptation for Medical Image Segmentation via Contrastive Retrieval-Augmented Joint Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5KRCUXWV}},
  note         = {Machine review of arXiv:2510.09953}
}
abstract

Manual medical image segmentation by clinicians, though accurate, is time-consuming and variable across experts, whereas AI-based models automate this process but often underperform with limited data and domain shifts. Inspired by how pathology trainees acquire disease recognition skills through guided comparison with expert-annotated slides and histopathology atlas reference images, we propose Joint Retrieval-Augmented Segmentation (J-RAS). This framework enables segmentation networks to learn with guidance. J-RAS jointly optimizes a segmentation model and a retrieval model through alternating contrastive and supervised learning, allowing the retrieval network to discover contextually relevant image-mask pairs that refine the segmentation model's anatomical reasoning. Unlike conventional retrieval-based augmentation that passively provides similar samples, J-RAS establishes a mutual adaptation and optimization loop where the retrieval model learns to emphasize segmentation-relevant cues, while the segmentation model leverages retrieved examples to improve boundary delineation, robustness to rare cases, and cross-dataset generalization. Evaluations on four public benchmarks spanning different imaging modalities, including ACDC and M&Ms (MRI), Breast Cancer Ultrasound, and lung and infection CT, across multiple backbones (U-Net, TransUNet, SAM, and SegFormer) demonstrate the generalizability and effectiveness of J-RAS. For instance, on ACDC, SegFormer improves from a mean Dice of 0.8708$\pm$0.042 and HD of 1.8130$\pm$2.49 to 0.9115$\pm$0.031 and 1.1489$\pm$0.30. These results highlight how retrieval-guided contrastive optimization bridges human-like guidance and machine-learned precision in medical image segmentation.

Figures

Figures reproduced from arXiv: 2510.09953 by the authors.

Figure 1
Figure 1. Overview of the proposed Joint Retrieval-Augmented Segmentation (J-RAS) method. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Mean Dice scores on the ACDC test set across di [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Dice scores for Classes (RV, MYO, and LV) using [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Case-level analysis of improvements and degradations on the [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 6
Figure 6. Figure 6: Qualitative segmentation results on the ACDC test set using [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Comparison of segmentation results on samples from the [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Comparison of segmentation results using [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 10
Figure 10. Figure 10: Comparison of fusion strategies in the J-RAS method on the ACDC test set. Results are reported using Dice score ( [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 12
Figure 12. Figure 12: Mean Dice scores on the M&Ms dataset using the J-RAS [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 8 linked inside Pith

  1. [43]

    L. Zhao, X. Chen, E. Z. Chen, Y . Liu, T. Chen, S. Sun, Retrieval-augmented few-shot medical im- age segmentation with foundation models, IEEE Transactions on Neural Networks and Learning Sys- tems (2025)

    2025

  2. [1]

    Minaee, Y

    S. Minaee, Y . Boykov, F. Porikli, A. Plaza, N. Ke- htarnavaz, D. Terzopoulos, Image segmentation us- ing deep learning: A survey, IEEE transactions on pattern analysis and machine intelligence 44 (7) (2021) 3523–3542

    2021

  3. [2]

    Y . Xu, R. Quan, W. Xu, Y . Huang, X. Chen, F. Liu, Advances in medical image segmentation: A com- prehensive review of traditional, deep learning and hybrid approaches, Bioengineering 11 (10) (2024) 1034

    2024

  4. [3]

    J. Li, P. Jiang, Q. An, G.-G. Wang, H.-F. Kong, Med- ical image identification methods: A review, Com- puters in Biology and Medicine 169 (2024) 107777

    2024

  5. [4]

    C. H. Sudre, Image segmentation, in: Advances in Magnetic Resonance Technology and Applications, V ol. 4, Elsevier, 2021, pp. 95–105

    2021

  6. [5]

    S. Wang, C. Li, R. Wang, Z. Liu, M. Wang, H. Tan, Y . Wu, X. Liu, H. Sun, R. Yang, et al., Annotation- efficient deep learning for automatic medical image segmentation, Nature communications 12 (1) (2021) 5915

    2021

  7. [6]

    Y . Gao, X. Chen, Q. Yang, A. Lasso, I. Kolesov, S. Pieper, R. Kikinis, A. Tannenbaum, L. Zhu, An effective and open source interactive 3d medical im- age segmentation solution, Scientific Reports 14 (1) (2024) 1–19

    2024

  8. [7]

    X. Liu, Z. Deng, Y . Yang, Recent progress in seman- tic image segmentation, Artificial Intelligence Re- view 52 (2) (2019) 1089–1106. 13

    2019

Show all 49 references
  1. [8]

    LeCun, Y

    Y . LeCun, Y . Bengio, G. Hinton, Deep learning, na- ture 521 (7553) (2015) 436–444

    2015

  2. [9]

    O’shea, R

    K. O’shea, R. Nash, An introduction to con- volutional neural networks, arXiv preprint arXiv:1511.08458 (2015)

    2015 arXiv

  3. [10]

    Dosovitskiy, L

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weis- senborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, et al., An image is worth 16x16 words: Transform- ers for image recognition at scale, arXiv preprint arXiv:2010.11929 (2020)

    2010 arXiv

  4. [11]

    Litjens, T

    G. Litjens, T. Kooi, B. E. Bejnordi, A. A. A. Se- tio, F. Ciompi, M. Ghafoorian, J. A. Van Der Laak, B. Van Ginneken, C. I. Sánchez, A survey on deep learning in medical image analysis, Medical image analysis 42 (2017) 60–88

    2017

  5. [12]

    Ghafoorian, A

    M. Ghafoorian, A. Mehrtash, T. Kapur, N. Karsse- meijer, E. Marchiori, M. Pesteie, C. R. Guttmann, F.-E. De Leeuw, C. M. Tempany, B. Van Ginneken, et al., Transfer learning for domain adaptation in mri: Application in brain lesion segmentation, in: International conference on ...

    2017

  6. [13]

    Dimitrovski, D

    I. Dimitrovski, D. Kocev, S. Loskovska, S. Džeroski, Hierarchical annotation of medical images, Pattern Recognition 44 (10-11) (2011) 2436–2449

    2011

  7. [14]

    C. B. Akgül, D. L. Rubin, S. Napel, C. F. Beaulieu, H. Greenspan, B. Acar, Content-based image re- trieval in radiology: current status and future direc- tions, Journal of digital imaging 24 (2) (2011) 208– 222

    2011

  8. [15]

    T. M. Lehmann, M. O. Güld, C. Thies, B. Fis- cher, K. Spitzer, D. Keysers, H. Ney, M. Kohnen, H. Schubert, B. B. Wein, Content-based image re- trieval in medical applications, Methods of informa- tion in medicine 43 (04) (2004) 354–361

    2004

  9. [16]

    Ronneberger, P

    O. Ronneberger, P. Fischer, T. Brox, U-net: Con- volutional networks for biomedical image segmen- tation, in: Medical image computing and computer- assisted intervention–MICCAI 2015: 18th interna- tional conference, Munich, Germany, October 5-9, 2015, proceedings, part III 18,...

    2015

  10. [17]

    J. Chen, Y . Lu, Q. Yu, X. Luo, E. Adeli, Y . Wang, L. Lu, A. L. Yuille, Y . Zhou, Transunet: Transform- ers make strong encoders for medical image seg- mentation, arXiv preprint arXiv:2102.04306 (2021)

    2021 arXiv

  11. [18]

    E. Xie, W. Wang, Z. Yu, A. Anandkumar, J. M. Al- varez, P. Luo, Segformer: Simple and efficient de- sign for semantic segmentation with transformers, Advances in neural information processing systems 34 (2021) 12077–12090

    2021

  12. [19]

    Kirillov, E

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y . Lo, et al., Segment anything, in: Proceedings of the IEEE/CVF international conference on com- puter vision, 2023, pp. 4015–4026

    2023

  13. [20]

    Bernard, A

    O. Bernard, A. Lalande, C. Zotti, F. Cervenan- sky, X. Yang, P.-A. Heng, I. Cetin, K. Lekadir, O. Camara, M. A. G. Ballester, et al., Deep learn- ing techniques for automatic mri cardiac multi- structures segmentation and diagnosis: is the prob- lem solved?, IEEE transactions ...

    2018

  14. [21]

    V . M. Campello, P. Gkontra, C. Izquierdo, C. Martin-Isla, A. Sojoudi, P. M. Full, K. Maier- Hein, Y . Zhang, Z. He, J. Ma, et al., Multi-centre, multi-vendor and multi-disease cardiac segmenta- tion: the m&ms challenge, IEEE Transactions on Medical Imaging 40 (12) (2021) 3543–3554

    2021

  15. [22]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, I. Polosukhin, At- tention is all you need, Advances in neural informa- tion processing systems 30 (2017)

    2017

  16. [23]

    H. Cao, Y . Wang, J. Chen, D. Jiang, X. Zhang, Q. Tian, M. Wang, Swin-unet: Unet-like pure trans- former for medical image segmentation, in: Eu- ropean conference on computer vision, Springer, 2022, pp. 205–218

    2022

  17. [24]

    Z. Wang, D. He, L. Zhao, B. Liu, Y . Zheng, X. Zhang, Difusionseg: Diffusion-driven seman- tic segmentation with multi-modal image fusion for 14 enhanced perception, Knowledge-Based Systems (2025) 114481

    2025

  18. [25]

    Zubair, M

    M. Zubair, M. Hussai, M. A. Al-Bashrawi, M. Ben- dechache, M. Owais, A comprehensive review of techniques, algorithms, advancements, challenges, and clinical applications of multi-modal medical im- age fusion for improved diagnosis, arXiv preprint arXiv:2505.14715 (2025)

    2025 arXiv

  19. [26]

    Z. Guo, X. Li, H. Huang, N. Guo, Q. Li, Deep learning-based image segmentation on multimodal medical imaging, IEEE transactions on radiation and plasma medical sciences 3 (2) (2019) 162–169

    2019

  20. [27]

    Albekairi, M

    M. Albekairi, M. v. O. Mohamed, K. Kaaniche, G. Abbas, M. D. Alanazi, T. M. Alanazi, A. Emara, Multimodal medical image fusion com- bining saliency perception and generative adversar- ial network, Scientific Reports 15 (1) (2025) 10609

    2025

  21. [28]

    J. Ren, J. G. Eriksen, J. Nijkamp, S. S. Korre- man, Comparing different ct, pet and mri multi- modality image combinations for deep learning- based head and neck tumor segmentation, Acta On- cologica 60 (11) (2021) 1399–1406

    2021

  22. [29]

    B. J. Pichler, H. F. Wehrl, A. Kolb, M. S. Juden- hofer, Positron emission tomography/magnetic reso- nance imaging: the next generation of multimodality imaging?, in: Seminars in nuclear medicine, V ol. 38, Elsevier, 2008, pp. 199–208

    2008

  23. [30]

    K. F. Hossain, S. A. Kamran, J. Ong, A. Tavakkoli, Enhancing efficient deep learning models with mul- timodal, multi-teacher insights for medical im- age segmentation, Scientific Reports 15 (1) (2025) 15948

    2025

  24. [31]

    Jafari, S

    M. Jafari, S. Francis, J. M. Garibaldi, X. Chen, Lmisa: A lightweight multi-modality image seg- mentation network via domain adaptation using gra- dient magnitude and shape constraint, Medical Im- age Analysis 81 (2022) 102536

    2022

  25. [32]

    C. Liao, K. Lei, X. Zheng, J. Moon, Z. Wang, Y . Wang, D. P. Paudel, L. Van Gool, X. Hu, Bench- marking multi-modal semantic segmentation under sensor failures: Missing and noisy modality robust- ness, in: Proceedings of the Computer Vision and Pattern Recognition Conference, ...

    2025

  26. [33]

    J. Tan, X. Zheng, Y . Liu, Robust multimodal segmentation with representation regularization and hybrid prototype distillation, arXiv preprint arXiv:2505.12861 (2025)

    2025 arXiv

  27. [34]

    J. E. Iglesias, M. R. Sabuncu, Multi-atlas segmenta- tion of biomedical images: a survey, Medical image analysis 24 (1) (2015) 205–219

    2015

  28. [35]

    H. Wang, J. W. Suh, S. R. Das, J. B. Pluta, C. Craige, P. A. Yushkevich, Multi-atlas segmentation with joint label fusion, IEEE transactions on pattern anal- ysis and machine intelligence 35 (3) (2012) 611– 623

    2012

  29. [36]

    R. A. Heckemann, J. V . Hajnal, P. Aljabar, D. Rueckert, A. Hammers, Automatic anatomical brain mri segmentation combining label propagation and decision fusion, NeuroImage 33 (1) (2006) 115– 126

    2006

  30. [37]

    M. R. Sabuncu, B. T. Yeo, K. Van Leemput, B. Fis- chl, P. Golland, A generative model for image seg- mentation based on label fusion, IEEE transactions on medical imaging 29 (10) (2010) 1714–1729

    2010

  31. [38]

    X. Tang, K. Oishi, A. V . Faria, A. E. Hillis, M. S. Albert, S. Mori, M. I. Miller, Bayesian parameter es- timation and segmentation in the multi-atlas random orbit model, PloS one 8 (6) (2013) e65591

    2013

  32. [39]

    Aljabar, R

    P. Aljabar, R. A. Heckemann, A. Hammers, J. V . Ha- jnal, D. Rueckert, Multi-atlas based segmentation of brain images: atlas selection and its effect on accu- racy, Neuroimage 46 (3) (2009) 726–738

    2009

  33. [40]

    Shaban, S

    A. Shaban, S. Bansal, Z. Liu, I. Essa, B. Boots, One-shot learning for semantic segmentation, arXiv preprint arXiv:1709.03410 (2017)

    2017 arXiv

  34. [41]

    K. Wang, J. H. Liew, Y . Zou, D. Zhou, J. Feng, Panet: Few-shot image semantic segmentation with prototype alignment, in: proceedings of the IEEE/CVF international conference on computer vi- sion, 2019, pp. 9197–9206. 15

    2019

  35. [42]

    Z. Tian, H. Zhao, M. Shu, Z. Yang, R. Li, J. Jia, Prior guided feature enrichment network for few-shot seg- mentation, IEEE transactions on pattern analysis and machine intelligence 44 (2) (2020) 1050–1065

    2020

  36. [44]

    N. Ravi, V . Gabeur, Y .-T. Hu, R. Hu, C. Ryali, T. Ma, H. Khedr, R. Rädle, C. Rolland, L. Gustafson, et al., Sam 2: Segment anything in images and videos, arXiv preprint arXiv:2408.00714 (2024)

    2024 arXiv

  37. [45]

    T. Chen, S. Kornblith, M. Norouzi, G. Hinton, A simple framework for contrastive learning of visual representations, in: International conference on ma- chine learning, PmLR, 2020, pp. 1597–1607

    2020

  38. [46]

    Oquab, T

    M. Oquab, T. Darcet, T. Moutakanni, H. V o, M. Szafraniec, V . Khalidov, P. Fernandez, D. Haz- iza, F. Massa, A. El-Nouby, et al., Dinov2: Learn- ing robust visual features without supervision, arXiv preprint arXiv:2304.07193 (2023)

    2023 arXiv

  39. [47]

    Pérez-García, H

    F. Pérez-García, H. Sharma, S. Bond-Taylor, K. Bouzid, V . Salvatelli, M. Ilse, S. Bannur, D. C. Castro, A. Schwaighofer, M. P. Lungren, et al., Exploring scalable medical image encoders be- yond text supervision, Nature Machine Intelligence (2025) 1–12

    2025

  40. [48]

    L. R. Dice, Measures of the amount of ecologic as- sociation between species, Ecology 26 (3) (1945) 297–302

    1945

  41. [49]

    D. P. Huttenlocher, G. A. Klanderman, W. J. Ruck- lidge, Comparing images using the hausdorffdis- tance, IEEE Transactions on pattern analysis and machine intelligence 15 (9) (2002) 850–863. 16

    2002

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.