Pith. sign in

REVIEW 4 major objections 6 minor 79 references

Mamba-Sea: A Mamba-based Framework with Global-to-Local Sequence Augmentation for Generalizable Medical Image Segmentation

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read Mamba-Sea, a Mamba-based framework with global-to-local sequence augmentation, achieves state-of-the-art domain-generalized medical image segmentation, including the first Prostate Dice score above 90%.

desk verdict Solid incremental engineering with real benchmark gains, but the missing comparison to the same-group START paper undermines the first/SOTA claim. read the letter →

arxiv 2504.17515 v1 pith:2QMMDSCJ submitted 2025-04-24 cs.CV

classification cs.CV
keywords domaingeneralizationmedicalimagesegmentationstatespacemodelsMambasequenceaugmentationstyletransformationsemanticconsistencytrainingcross-domainrobustness
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 is trying to establish that Mamba, a selective state space model, can be turned into a strong backbone for domain-generalized medical image segmentation rather than remaining limited to supervised same-distribution tasks. It proposes Mamba-Sea, which augments images globally and token sequences locally during training, and enforces semantic consistency between original and augmented views. On leave-one-domain-out benchmarks the framework is reported to reach 90.02% average Dice on the Prostate dataset, 89.68% on the Fundus dataset, and 93.11% on a large skin-lesion dataset, each above the previous state of the art. If the claim holds, clinically deployed segmentation models could be made more robust to new scanners and protocols without quadratic attention cost.

What carries the argument

The load-bearing machinery is the local sequence-wise style transformation (LSA) placed inside the Mamba scanning blocks, coordinated with the global appearance variation augmentation (GVA) and a semantic consistency loss. LSA computes per-instance channel-wise means and standard deviations, estimates their uncertainty across the batch, resamples them from a Gaussian model, and mixes the transformed features with the original features using a random continuous mask over tokens (Eqs. 8-13). This is designed to perturb exactly the regions of the input-dependent state matrices where, according to the paper's hypothesis, domain-specific information accumulates, while the continuous-mask choice preserves local semantic structure. GVA supplies whole-image appearance diversity through a lightweight gating network, and the consistency loss makes the two augmentation levels work together.

What would settle it

Run the released implementation on the Prostate leave-one-domain-out protocol with the Gaussian resampling in Eq. (10) replaced by uniform or bootstrap resampling of the same per-batch statistics; if average Dice does not drop below 90%, the Gaussian modeling is not what carries the improvement.

Watch

Extended reading notes

Core claim

The paper's central claim is that the input-dependent matrices of Mamba's selective scanning mechanism accumulate domain-specific information during training, and that this accumulated bias is the main obstacle to cross-domain generalization. Mamba-Sea therefore inserts a local sequence-wise style transformation (LSA) into the scanning blocks, which models per-batch channel statistics as Gaussian, resamples them to synthesize new style shifts, and applies them only to random continuous subsequences of tokens via a mask. A global appearance variation augmentation (GVA) uses a learnable gating network to simulate site-to-site illumination differences over entire images, and a semantic consistency loss forces predictions on original and augmented inputs to match. The paper reports that this combined recipe yields state-of-the-art averages of 90.02% Dice on Prostate, 89.68% on Fundus, and 93.11% on skin lesion segmentation, with GVA and LSA discarded at inference time so that test-time cost is unchanged.

Load-bearing premise

The load-bearing premise is that per-batch feature statistics (channel-wise means and standard deviations) are distributed close enough to Gaussian that resampling them, masked over random continuous subsequences, produces realistic unseen-domain style shifts without damaging semantic content.

Editorial extensions

If this is right

  • Mamba-based models can serve as competitive backbones for domain-generalized medical segmentation, matching or surpassing CNN- and ViT-based DG methods without quadratic self-attention cost.
  • The per-batch Gaussian resampling of feature statistics and continuous token masking give a concrete, reusable recipe for style augmentation inside state space models.
  • Because GVA and LSA are training-only and discarded at inference, the framework adds negligible test-time latency and parameter overhead, easing clinical deployment.
  • Removing either augmentation module or the consistency loss measurably lowers average Dice on both benchmarks, so the gains depend on the combination rather than on the Mamba backbone alone.

Reading between the lines

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

  • Editorial inference: the Gaussian assumption on feature statistics is likely not the only source of the gains; comparing it against bootstrap or uniform resampling of the same statistics would isolate how much the distributional model matters.
  • Editorial inference: the GVA brightness threshold is hand-set per dataset (0.4 for Fundus, 0.03 for Prostate), so a learned or data-dependent threshold is a natural extension the paper leaves open.
  • Editorial inference: the reported 32% reduction in average feature-space domain gap in the decoder is a quantitative handle that could be applied to other SSM backbones to test whether the augmentation genuinely removes domain-specific accumulation.
  • Editorial inference: if the input-dependent matrices are indeed the carrier of domain bias, architectural changes to the state update itself, rather than augmentation only, should further improve generalization.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes Mamba-Sea, a domain-generalization framework for medical image segmentation built on a Mamba/VM-UNet backbone. It combines a global appearance variation augmentation (GVA) with a local sequence-wise style transformation (LSA) that resamples channel-wise feature statistics under a Gaussian assumption over a random continuous token mask, plus a semantic consistency loss between original and globally augmented views. The method is evaluated on Fundus, Prostate, and skin-lesion benchmarks under leave-one-domain-out protocols, reporting average Dice of 89.68%, 90.02%, and 93.11%, respectively. The paper claims to be the first Mamba-based DG framework for medical image segmentation and the first to exceed 90% average Dice on the Prostate dataset, with code released.

Significance. If the empirical results hold, Mamba-Sea would be a useful extension of Mamba to domain generalization, with a small computational overhead (GVA adds 2.58e-4 M parameters and training time increases by about 0.004 s per iteration), public code, and reported paired t-tests on all three datasets. The ablations and comparisons against CNN, ViT, and SAM-based baselines are extensive. However, the headline novelty and SOTA claims are not yet established because the cited prior state-space-model DG method START [36] is not included in any comparison table, and the hyperparameter selection protocol is unspecified. The Gaussian resampling assumption is a modeling choice; the paper's empirical value would remain if LSA were positioned as a stochastic style perturbation regularizer rather than a faithful model of unseen-domain statistics.

major comments (4)
  1. [Sec. IV.C; Tables II, III, IX] The manuscript claims to be the first Mamba-based DG framework for medical image segmentation and to achieve SOTA, yet it never compares against START [36], which is cited as the source of the LSA motivation (Sec. III.C.3) and of the domain-gap evaluation protocol (Sec. IV.C.3). Because [36] is titled 'A generalized state space model with saliency-driven token-aware transformation' and is prior work from the same group, it appears to be a directly relevant SSM-based DG method. Without a quantitative comparison on Fundus, Prostate, or skin lesion, the 'first' and SOTA claims are not established. Please add START to all comparison tables or explain why it is not a comparable prior method, and revise the novelty statements accordingly.
  2. [Sec. IV.B and IV.C.4] Hyperparameters are reported as fixed after sensitivity analysis: tau differs between Fundus (0.4) and Prostate (0.03), p is set to 0.75, and lambda to 0.1. The paper does not state whether these values were chosen by validation on held-out source domains or by observing target-domain test performance. In a DG evaluation, tuning on the target test set can inflate results and violates the held-out protocol. Furthermore, tau is not included in the sensitivity analysis of Table VII, so its effect is not quantified. Please specify the exact validation procedure used to select tau, p, and lambda, and if target-domain information was used, rerun the comparisons with source-only hyperparameter selection.
  3. [Sec. IV.C, Tables II-III, VIII] All Fundus and Prostate results are reported as means over three runs with no standard deviations, so the 1.41% Prostate margin over DCAC cannot be assessed against run-to-run variability. The paired t-tests in Table VIII are computed at the sample level; for Prostate, 2D slices from the same 3D volume are not independent, which can make sample-level p-values artificially small. Please report mean plus/minus standard deviation over runs, provide run-level comparisons, and use volume-level or clustered significance tests for Prostate.
  4. [Abstract and Table III] The claim 'first to surpass a Dice coefficient of 90% on the Prostate dataset' is ambiguous: the 90.02% is the average over six leave-one-out tasks, while the baseline VM-Unet already exceeds 90% on Domain 1 (91.02) and Domain 6 (90.91), and DCAC reaches 91.76 on Domain 1. Please state explicitly that the record is the average across domains, and compare against the prior best average of 88.61% as given in the text.
minor comments (6)
  1. [Sec. III.C.2, Eq. (5)-(7)] The notation f_g(x) is used both for the input features to GVA and for the transformed output in Eq. (7), which is confusing; please use separate symbols for the input and the augmented output.
  2. [Table IX] Several reference numbers in Table IX are incorrect: WT-PSE is [62] not [37], SAM-Med2D is [63] not [62], and Med-SA is [7] not [63]; please correct the citations.
  3. [Sec. IV.C.3] The t-SNE and MMD analyses measure feature alignment, which may partly reflect the effect of the consistency loss rather than genuine domain-invariant semantic learning; consider an additional analysis that separates alignment from semantic content preservation.
  4. [Sec. III.C.4 and IV.B] The architecture of the GVA lightweight network is not described in enough detail (number of enhancement blocks, channel widths, activation functions), and the skin lesion experiment does not state whether the same training configuration and hyperparameters as Fundus/Prostate were used; please provide these reproducibility details.
  5. [Sec. III.C.3] The choice of continuous sub-sequence perturbation over randomly scattered token perturbation is asserted to preserve token dependencies, but no ablation compares the continuous mask in Eq. (12) with an i.i.d. random mask of the same proportion; please add such an ablation or soften the claim.
  6. [Throughout] There are several typos and reference inconsistencies, including 'pervious' for 'previous', 'matrics' for 'matrices', and the duplicated LKM-UNet description in Sec. II.B; a careful proofreading pass is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the core SOTA claims rest on held-out benchmark evaluations and are not derived from the method's own fitted quantities.

full rationale

I walked the paper's derivation chain and found no step in which a prediction is equivalent to an input by construction, a fitted parameter is renamed as a prediction, or a load-bearing result reduces to an unverified self-citation. The central claims (90.02% Dice on Prostate, 89.68% on Fundus, 93.11% on skin lesion) are measured on unseen target domains after leave-one-domain-out training, which is an external, held-out evaluation rather than an objective function being optimized. The GVA and LSA augmentations are modeling choices: LSA's Gaussian-resampling assumption (Eqs. 8-12) is a hypothesis about feature statistics, and even if the Gaussian model is imperfect, LSA could still act as a regularizer; this does not make the reported Dice a fitted target. The semantic consistency loss (Eq. 16) enforces agreement between predictions, but the quality of those predictions is judged against ground-truth labels, so the evaluation is not self-confirmatory. The t-SNE and domain-gap analyses visualize the model's own features, but they are supplementary evidence, not the basis of the headline SOTA claim. Reference [36] (START) is a self-citation used to motivate the design and the domain-gap evaluation protocol, but the paper does not derive its quantitative gains from [36]; the gains are established by direct comparison with baselines and prior methods. The omission of START from Tables II, III, and IX is a potential completeness/novelty concern, not circularity, because no equation or result in this paper is shown to reduce to START's outputs. Overall, the derivation is self-contained against external benchmarks, so the appropriate finding is no significant circularity.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claim rests on standard supervised segmentation and style-augmentation assumptions rather than new physics or entities. The main extra degrees of freedom are the three hyperparameters tau, p, and lambda, tuned per dataset; no new entities are introduced. The Gaussian-statistics and continuous-mask assumptions in LSA are design choices with limited standalone validation.

free parameters (3)
  • GVA gating threshold tau = 0.4 (Fundus), 0.03 (Prostate)
    Controls which low-brightness images receive global augmentation (Eq. 6); set differently per dataset with sensitivity not reported.
  • LSA mask proportion p = 0.75
    Fraction of tokens in random continuous sub-sequences whose style is resampled (Eq. 12); chosen by sensitivity analysis in Table VII.
  • Consistency loss weight lambda = 0.1
    Balances segmentation and consistency losses (Eq. 17); sensitivity in Table VII shows mild impact.
assumptions (6)
  • domain assumption Mamba's input-dependent matrices accumulate domain-specific information during training, which harms cross-domain generalization.
    Stated in Section I and attributed to reference [36]; it motivates LSA but is not re-derived in this paper.
  • domain assumption Channel-wise mean and standard deviation of features encode domain-specific style, and perturbing them simulates unseen domains.
    Standard style-augmentation assumption from MixStyle [52] and ALOFT [53]; used in Eqs. (8)-(11).
  • ad hoc to paper A random continuous sub-sequence mask preserves token dependencies better than isolated token perturbations.
    Introduced in Section III.C.3 without empirical comparison to isolated-token masking.
  • ad hoc to paper Feature statistics follow a multivariate Gaussian distribution with means and variances estimated per batch.
    Eq. (10); a modeling assumption for uncertainty estimation, not validated against actual feature statistics.
  • domain assumption Globally augmented images share the same segmentation label as original images.
    Used in the consistency loss Eq. (16); standard for appearance augmentation.
  • standard math The Mamba ODE discretization in Eqs. (1)-(4) correctly models the VM-UNet backbone.
    Background from S4 and Mamba, accepted from prior literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mamba-Sea: A Mamba-based Framework with Global-to-Local Sequence Augmentation for Generalizable Medical Image Segmentation." pith.science (2026). https://pith.science/paper/2QMMDSCJ

@misc{pith2026250417515,
  author       = {Pith},
  title        = {Pith review of: Mamba-Sea: A Mamba-based Framework with Global-to-Local Sequence Augmentation for Generalizable Medical Image Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2QMMDSCJ}},
  note         = {Machine review of arXiv:2504.17515}
}
read the original abstract

To segment medical images with distribution shifts, domain generalization (DG) has emerged as a promising setting to train models on source domains that can generalize to unseen target domains. Existing DG methods are mainly based on CNN or ViT architectures. Recently, advanced state space models, represented by Mamba, have shown promising results in various supervised medical image segmentation. The success of Mamba is primarily owing to its ability to capture long-range dependencies while keeping linear complexity with input sequence length, making it a promising alternative to CNNs and ViTs. Inspired by the success, in the paper, we explore the potential of the Mamba architecture to address distribution shifts in DG for medical image segmentation. Specifically, we propose a novel Mamba-based framework, Mamba-Sea, incorporating global-to-local sequence augmentation to improve the model's generalizability under domain shift issues. Our Mamba-Sea introduces a global augmentation mechanism designed to simulate potential variations in appearance across different sites, aiming to suppress the model's learning of domain-specific information. At the local level, we propose a sequence-wise augmentation along input sequences, which perturbs the style of tokens within random continuous sub-sequences by modeling and resampling style statistics associated with domain shifts. To our best knowledge, Mamba-Sea is the first work to explore the generalization of Mamba for medical image segmentation, providing an advanced and promising Mamba-based architecture with strong robustness to domain shifts. Remarkably, our proposed method is the first to surpass a Dice coefficient of 90% on the Prostate dataset, which exceeds previous SOTA of 88.61%. The code is available at https://github.com/orange-czh/Mamba-Sea.

Figures

Figures reproduced from arXiv: 2504.17515 by the authors.

Figure 1
Figure 1. Typical images from three public medical datasets with [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall structure of Mamba-Sea. (a) represents [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Examples of original images alongside their augmented [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Structure of LSA. For each sequence, a unique mask is generated to enhance the diversity of the augmentation. detailed operation of LSA applied to the input is presented in [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Visualization on segmentation results of different methods on [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Visualization on segmentation results of different methods on [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: t-SNE visualization [66] of the features extracted from original images, after training with VM-UNet [8] and Mamba￾Sea on the Fundus dataset. Different colors represent different domains. The distance between domains is computed as the average Euclidean distance betwee…
Figure 8
Figure 8. Figure 8: Visualization on segmentation results of different methods on [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

79 extracted references · 48 canonical work pages

  1. [36]

    Start: A generalized state space model with saliency-driven token-aware transformation,

    J. Guo, L. Qi, Y . Shi, and Y . Gao, “Start: A generalized state space model with saliency-driven token-aware transformation,” in NeurIPS, 2024

  2. [1]

    U-net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in MICCAI. Springer, 2015, pp. 234–241

  3. [2]

    H-denseunet: hybrid densely connected unet for liver and tumor segmentation from ct volumes,

    X. Li, H. Chen, X. Qi, Q. Dou, C.-W. Fu, and P.-A. Heng, “H-denseunet: hybrid densely connected unet for liver and tumor segmentation from ct volumes,” IEEE TMI, vol. 37, no. 12, pp. 2663–2674, 2018

  4. [3]

    Segmentation of arm ultrasound images in breast cancer- related lymphedema: A database and deep learning algorithm,

    S. Goudarzi, J. Whyte, M. Boily, A. Towers, R. D. Kilgour, and H. Rivaz, “Segmentation of arm ultrasound images in breast cancer- related lymphedema: A database and deep learning algorithm,” IEEE TBME, vol. 70, no. 9, pp. 2552–2563, 2023

  5. [4]

    Shape-aware meta-learning for generalizing prostate mri segmentation to unseen domains,

    Q. Liu, Q. Dou, and P.-A. Heng, “Shape-aware meta-learning for generalizing prostate mri segmentation to unseen domains,” in MICCAI. Springer, 2020, pp. 475–485

  6. [5]

    Cddsa: Contrastive domain disentanglement and style augmentation for generalizable medical image segmentation,

    R. Gu, G. Wang, J. Lu, J. Zhang, W. Lei, Y . Chen, W. Liao, S. Zhang, K. Li, D. N. Metaxas et al., “Cddsa: Contrastive domain disentanglement and style augmentation for generalizable medical image segmentation,” MedIA, vol. 89, p. 102904, 2023

  7. [6]

    Swin-unet: Unet-like pure transformer for medical image segmenta- tion,

    H. Cao, Y . Wang, J. Chen, D. Jiang, X. Zhang, Q. Tian, and M. Wang, “Swin-unet: Unet-like pure transformer for medical image segmenta- tion,” in ECCV Workshop, 2022, pp. 205–218

  8. [7]

    Medical sam adapter: Adapting segment anything model for medical image segmentation,

    J. Wu, Z. Wang, M. Hong, W. Ji, H. Fu, Y . Xu, M. Xu, and Y . Jin, “Medical sam adapter: Adapting segment anything model for medical image segmentation,” MedIA, p. 103547, 2025

Show all 79 references
  1. [8]

    Vm-unet: Vision mamba unet for medical image segmentation,

    J. Ruan and S. Xiang, “Vm-unet: Vision mamba unet for medical image segmentation,” arXiv preprint arXiv:2402.02491 , 2024

  2. [9]

    Semi-supervised meta-learning with disentanglement for domain-generalised medical image segmentation,

    X. Liu, S. Thermos, A. O’Neil, and S. A. Tsaftaris, “Semi-supervised meta-learning with disentanglement for domain-generalised medical image segmentation,” in MICCAI. Springer, 2021, pp. 307–317

  3. [10]

    Structure- preserving color normalization and sparse stain separation for histolog- ical images,

    A. Vahadane, T. Peng, A. Sethi, S. Albarqouni, L. Wang, M. Baust, K. Steiger, A. M. Schlitter, I. Esposito, and N. Navab, “Structure- preserving color normalization and sparse stain separation for histolog- ical images,” IEEE TMI, vol. 35, no. 8, pp. 1962–1971, 2016

  4. [11]

    Domain gen- eralization in restoration of cataract fundus images via high-frequency components,

    H. Liu, H. Li, M. Ou, Y . Zhao, H. Qi, Y . Hu, and J. Liu, “Domain gen- eralization in restoration of cataract fundus images via high-frequency components,” in IEEE ISBI, 2022, pp. 1–5

  5. [12]

    Generalizable cross- modality medical image segmentation via style augmentation and dual normalization,

    Z. Zhou, L. Qi, X. Yang, D. Ni, and Y . Shi, “Generalizable cross- modality medical image segmentation via style augmentation and dual normalization,” in ECCV, 2022, pp. 20 856–20 865

  6. [13]

    Domain generalization for mammography detection via multi-style and multi-view contrastive learning,

    Z. Li, Z. Cui, S. Wang, Y . Qi, X. Ouyang, Q. Chen, Y . Yang, Z. Xue, D. Shen, and J.-Z. Cheng, “Domain generalization for mammography detection via multi-style and multi-view contrastive learning,” in MIC- CAI. Springer, 2021, pp. 98–108

  7. [14]

    Aadg: Automatic augmentation for domain generalization on retinal image segmentation,

    J. Lyu, Y . Zhang, Y . Huang, L. Lin, P. Cheng, and X. Tang, “Aadg: Automatic augmentation for domain generalization on retinal image segmentation,” IEEE TMI, vol. 41, no. 12, pp. 3699–3711, 2022

  8. [15]

    Improving vision transformers by revisiting high-frequency components,

    J. Bai, L. Yuan, S.-T. Xia, S. Yan, Z. Li, and W. Liu, “Improving vision transformers by revisiting high-frequency components,” in ECCV. Springer, 2022, pp. 1–18

  9. [16]

    Dimix: Disentangle-and-mix based do- main generalizable medical image segmentation,

    H. Kim, Y . Shin, and D. Hwang, “Dimix: Disentangle-and-mix based do- main generalizable medical image segmentation,” in MICCAI. Springer, 2023, pp. 242–251

  10. [17]

    Transformers in medical imaging: A survey,

    F. Shamshad, S. Khan, S. W. Zamir, M. H. Khan, M. Hayat, F. S. Khan, and H. Fu, “Transformers in medical imaging: A survey,”MedIA, vol. 88, p. 102802, 2023

  11. [18]

    Mamba: Linear-time sequence modeling with selective state spaces,

    A. Gu and T. Dao, “Mamba: Linear-time sequence modeling with selective state spaces,” arXiv preprint arXiv:2312.00752 , 2023

  12. [19]

    Log-vmamba: Local- global vision mamba for medical image segmentation,

    T. D. Q. Dang, H. H. Nguyen, and A. Tiulpin, “Log-vmamba: Local- global vision mamba for medical image segmentation,” in ACCV, 2024, pp. 548–565

  13. [20]

    Mambavision: A hybrid mamba- transformer vision backbone,

    A. Hatamizadeh and J. Kautz, “Mambavision: A hybrid mamba- transformer vision backbone,” in CVPR, 2025

  14. [21]

    Vssd: Vision mamba with non- causal state space duality,

    Y . Shi, M. Dong, M. Li, and C. Xu, “Vssd: Vision mamba with non- causal state space duality,” arXiv preprint arXiv:2407.18559 , 2024

  15. [22]

    Medmamba: Vision mamba for medical image classification,

    Y . Yue and Z. Li, “Medmamba: Vision mamba for medical image classification,” arXiv preprint arXiv:2403.03849 , 2024

  16. [23]

    Vision mamba for classification of breast ultrasound images,

    A. Nasiri-Sarvi, M. S. Hosseini, and H. Rivaz, “Vision mamba for classification of breast ultrasound images,” in MICCAI Deep-Breath Workshop. Springer, 2024, pp. 148–158

  17. [24]

    Mamba-unet: Unet- like pure visual mamba for medical image segmentation,

    Z. Wang, J.-Q. Zheng, Y . Zhang, G. Cui, and L. Li, “Mamba-unet: Unet- like pure visual mamba for medical image segmentation,” arXiv preprint arXiv:2402.05079, 2024

  18. [25]

    Polyp-mamba: Polyp segmentation with visual mamba,

    Z. Xu, F. Tang, Z. Chen, Z. Zhou, W. Wu, Y . Yang, Y . Liang, J. Jiang, X. Cai, and J. Su, “Polyp-mamba: Polyp segmentation with visual mamba,” in MICCAI. Springer, 2024, pp. 510–521

  19. [26]

    Segmamba: Long-range sequential modeling mamba for 3d medical image segmentation,

    Z. Xing, T. Ye, Y . Yang, G. Liu, and L. Zhu, “Segmamba: Long-range sequential modeling mamba for 3d medical image segmentation,” in MICCAI. Springer, 2024, pp. 578–588

  20. [27]

    Hc-mamba: Vision mamba with hybrid convolutional techniques for medical image segmentation,

    J. Xu, “Hc-mamba: Vision mamba with hybrid convolutional techniques for medical image segmentation,” arXiv preprint arXiv:2405.05007 , 2024

  21. [28]

    Ultralight vm-unet: Parallel vision mamba significantly reduces parameters for skin lesion segmen- tation,

    R. Wu, Y . Liu, P. Liang, and Q. Chang, “Ultralight vm-unet: Parallel vision mamba significantly reduces parameters for skin lesion segmen- tation,” arXiv preprint arXiv:2403.20035 , 2024

  22. [29]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y . Loet al., “Segment anything,” in ICCV, 2023, pp. 4015–4026

  23. [30]

    Lora: Low-rank adaptation of large language models,

    E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, and W. Chen, “Lora: Low-rank adaptation of large language models,” arXiv preprint arXiv:2106.09685 , 2021

  24. [31]

    Multi- centre, multi-vendor and multi-disease cardiac segmentation: the m&ms challenge,

    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 segmentation: the m&ms challenge,” IEEE TMI, vol. 40, no. 12, pp. 3543–3554, 2021

  25. [32]

    Dofe: Domain-oriented feature embedding for generalizable fundus image segmentation on unseen datasets,

    S. Wang, L. Yu, K. Li, X. Yang, C.-W. Fu, and P.-A. Heng, “Dofe: Domain-oriented feature embedding for generalizable fundus image segmentation on unseen datasets,” IEEE TMI, vol. 39, no. 12, pp. 4237– 4248, 2020

  26. [33]

    Minimax estimation of maximum mean discrepancy with radial kernels,

    I. O. Tolstikhin, B. K. Sriperumbudur, and B. Sch ¨olkopf, “Minimax estimation of maximum mean discrepancy with radial kernels,” NeurIPS, vol. 29, 2016

  27. [34]

    VMamba: Visual state space model,

    Y . Liu, Y . Tian, Y . Zhao, H. Yu, L. Xie, Y . Wang, Q. Ye, J. Jiao, and Y . Liu, “VMamba: Visual state space model,” in NeurIPS, 2024

  28. [35]

    U-mamba: Enhancing long-range dependency for biomedical image segmentation,

    J. Ma, F. Li, and B. Wang, “U-mamba: Enhancing long-range dependency for biomedical image segmentation,” arXiv preprint arXiv:2401.04722, 2024

  29. [37]

    Domain and content adaptive convolution based multi-source domain generalization for medical image segmentation,

    S. Hu, Z. Liao, J. Zhang, and Y . Xia, “Domain and content adaptive convolution based multi-source domain generalization for medical image segmentation,” IEEE TMI, vol. 42, no. 1, pp. 233–244, 2022

  30. [38]

    Feddg: Federated domain generalization on medical image segmentation via episodic learning in continuous frequency space,

    Q. Liu, C. Chen, J. Qin, Q. Dou, and P.-A. Heng, “Feddg: Federated domain generalization on medical image segmentation via episodic learning in continuous frequency space,” in CVPR, 2021, pp. 1013– 1023. 15

  31. [39]

    Do- main generalization via model-agnostic learning of semantic features,

    Q. Dou, D. Coelho de Castro, K. Kamnitsas, and B. Glocker, “Do- main generalization via model-agnostic learning of semantic features,” Advances in neural information processing systems , vol. 32, 2019

  32. [40]

    Random style transfer based domain generalization net- works integrating shape and spatial information,

    L. Li, V . A. Zimmer, W. Ding, F. Wu, L. Huang, J. A. Schnabel, and X. Zhuang, “Random style transfer based domain generalization net- works integrating shape and spatial information,” in MICCAI. Springer, 2021, pp. 208–218

  33. [41]

    Tesla: Test-time self-learning with automatic adversarial augmentation,

    D. Tomar, G. Vray, B. Bozorgtabar, and J.-P. Thiran, “Tesla: Test-time self-learning with automatic adversarial augmentation,” in CVPR, 2023, pp. 20 341–20 350

  34. [42]

    Vm-unet-v2: Rethinking vision mamba unet for medical image segmentation,

    M. Zhang, Y . Yu, S. Jin, L. Gu, T. Ling, and X. Tao, “Vm-unet-v2: Rethinking vision mamba unet for medical image segmentation,” in International Symposium on Bioinformatics Research and Applications . Springer, 2024, pp. 335–346

  35. [43]

    Large window-based mamba unet for medical image segmentation: Beyond convolution and self- attention,

    J. Wang, J. Chen, D. Chen, and J. Wu, “Large window-based mamba unet for medical image segmentation: Beyond convolution and self- attention,” arXiv preprint arXiv:2403.07332 , 2024

  36. [44]

    Vim-unet: Vision mamba for biomedical segmentation,

    A. Archit and C. Pape, “Vim-unet: Vision mamba for biomedical segmentation,” arXiv preprint arXiv:2404.07705 , 2024

  37. [45]

    Con- volution and attention-free mamba-based cardiac image segmentation,

    A. Khan, M. Asad, M. Benning, C. Roney, and G. Slabaugh, “Con- volution and attention-free mamba-based cardiac image segmentation,” arXiv preprint arXiv:2406.05786 , 2024

  38. [46]

    Efficiently modeling long sequences with structured state spaces,

    A. Gu, K. Goel, and C. R ´e, “Efficiently modeling long sequences with structured state spaces,” arXiv preprint arXiv:2111.00396 , 2021

  39. [47]

    nnu-net: a self-configuring method for deep learning-based biomedical image segmentation,

    F. Isensee, P. F. Jaeger, S. A. Kohl, J. Petersen, and K. H. Maier-Hein, “nnu-net: a self-configuring method for deep learning-based biomedical image segmentation,” Nature methods, vol. 18, no. 2, pp. 203–211, 2021

  40. [48]

    Medsrgan: medical images super-resolution using generative adversarial networks,

    Y . Gu, Z. Zeng, H. Chen, J. Wei, Y . Zhang, B. Chen, Y . Li, Y . Qin, Q. Xie, Z. Jiang et al. , “Medsrgan: medical images super-resolution using generative adversarial networks,” Multimedia Tools and Applica- tions, vol. 79, pp. 21 815–21 840, 2020

  41. [49]

    On the proper use of structural similarity for the robust evaluation of medical image synthesis models,

    D. Gourdeau, S. Duchesne, and L. Archambault, “On the proper use of structural similarity for the robust evaluation of medical image synthesis models,” Medical Physics, vol. 49, no. 4, pp. 2462–2474, 2022

  42. [50]

    Toward fast, flexible, and robust low-light image enhancement,

    L. Ma, T. Ma, R. Liu, X. Fan, and Z. Luo, “Toward fast, flexible, and robust low-light image enhancement,” in CVPR, 2022, pp. 5637–5646

  43. [51]

    Vision mamba: Efficient visual representation learning with bidirectional state space model,

    L. Zhu, B. Liao, Q. Zhang, X. Wang, W. Liu, and X. Wang, “Vision mamba: Efficient visual representation learning with bidirectional state space model,” in ICML, 2024

  44. [52]

    Domain generalization with mixstyle,

    K. Zhou, Y . Yang, Y . Qiao, and T. Xiang, “Domain generalization with mixstyle,” in ICLR, 2021

  45. [53]

    Aloft: A lightweight mlp- like architecture with dynamic low-frequency transform for domain generalization,

    J. Guo, N. Wang, L. Qi, and Y . Shi, “Aloft: A lightweight mlp- like architecture with dynamic low-frequency transform for domain generalization,” in CVPR, 2023, pp. 24 132–24 141

  46. [54]

    A comprehensive retinal image dataset for the assessment of glaucoma from the optic nerve head analysis,

    J. Sivaswamy, S. Krishnadas, A. Chakravarty, G. Joshi, A. S. Tabish et al. , “A comprehensive retinal image dataset for the assessment of glaucoma from the optic nerve head analysis,” JSM Biomedical Imaging Data Papers, vol. 2, no. 1, p. 1004, 2015

  47. [55]

    Rim-one: An open retinal image database for optic nerve evaluation,

    F. Fumero, S. Alay ´on, J. L. Sanchez, J. Sigut, and M. Gonzalez- Hernandez, “Rim-one: An open retinal image database for optic nerve evaluation,” in CBMS. IEEE, 2011, pp. 1–6

  48. [56]

    Refuge challenge: A unified framework for evaluating automated methods for glaucoma assessment from fundus photographs,

    J. I. Orlando, H. Fu, J. B. Breda, K. Van Keer, D. R. Bathula, A. Diaz- Pinto, R. Fang, P.-A. Heng, J. Kim, J. Lee et al. , “Refuge challenge: A unified framework for evaluating automated methods for glaucoma assessment from fundus photographs,” MedIA, vol. 59, p. 101570, 2020

  49. [57]

    Nci- isbi 2013 challenge: automated segmentation of prostate structures,

    N. Bloch, A. Madabhushi, H. Huisman, J. Freymann, J. Kirby, M. Grauer, A. Enquobahrie, C. Jaffe, L. Clarke, and K. Farahani, “Nci- isbi 2013 challenge: automated segmentation of prostate structures,” The Cancer Imaging Archive , vol. 370, 2015

  50. [58]

    Computer-aided detection and diagnosis for prostate cancer based on mono and multi-parametric mri: a review,

    G. Lema ˆıtre, R. Mart ´ı, J. Freixenet, J. C. Vilanova, P. M. Walker, and F. Meriaudeau, “Computer-aided detection and diagnosis for prostate cancer based on mono and multi-parametric mri: a review,” CBM, vol. 60, pp. 8–31, 2015

  51. [59]

    Evaluation of prostate segmentation algorithms for mri: the promise12 challenge,

    G. Litjens, R. Toth, W. van de Ven, C. Hoeks, S. Kerkstra, B. van Gin- neken, G. Vincent, G. Guillard, N. Birbeck, J. Zhang et al., “Evaluation of prostate segmentation algorithms for mri: the promise12 challenge,” MedIA, vol. 18, no. 2, pp. 359–373, 2014

  52. [60]

    Domain generalization by solving jigsaw puzzles,

    F. M. Carlucci, A. D’Innocente, S. Bucci, B. Caputo, and T. Tommasi, “Domain generalization by solving jigsaw puzzles,” in CVPR, 2019, pp. 2229–2238

  53. [61]

    Generalizing deep learning for medical image segmentation to unseen domains via deep stacked transformation,

    L. Zhang, X. Wang, D. Yang, T. Sanford, S. Harmon, B. Turkbey, B. J. Wood, H. Roth, A. Myronenko, D. Xuet al., “Generalizing deep learning for medical image segmentation to unseen domains via deep stacked transformation,” IEEE TMI, vol. 39, no. 7, pp. 2531–2540, 2020

  54. [62]

    Learning robust shape regularization for generalizable medical image segmentation,

    K. Chen, T. Qin, V . H.-F. Lee, H. Yan, and H. Li, “Learning robust shape regularization for generalizable medical image segmentation,” IEEE TMI, 2024

  55. [63]

    Sam-med2d,

    J. Cheng, J. Ye, Z. Deng, J. Chen, T. Li, H. Wang, Y . Su, Z. Huang, J. Chen, L. Jiang et al., “Sam-med2d,” arXiv preprint arXiv:2308.16184, 2023

  56. [64]

    Imagenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Imagenet: A large-scale hierarchical image database,” inCVPR, 2009, pp. 248–255

  57. [65]

    Cutmix: Reg- ularization strategy to train strong classifiers with localizable features,

    S. Yun, D. Han, S. J. Oh, S. Chun, J. Choe, and Y . Yoo, “Cutmix: Reg- ularization strategy to train strong classifiers with localizable features,” in ICCV, 2019, pp. 6023–6032

  58. [66]

    Visualizing data using t-sne

    L. Van der Maaten and G. Hinton, “Visualizing data using t-sne.” JMLR, vol. 9, no. 11, 2008

  59. [67]

    Esp-medsam: Efficient self-prompting sam for universal domain-generalized medical image segmentation,

    Q. Xu, J. Li, X. He, Z. Liu, Z. Chen, W. Duan, C. Li, M. M. He, F. B. Tesema, W. P. Cheah et al., “Esp-medsam: Efficient self-prompting sam for universal domain-generalized medical image segmentation,” arXiv preprint arXiv:2407.14153, 2024

  60. [68]

    The ham10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions,

    P. Tschandl, C. Rosendahl, and H. Kittler, “The ham10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions,” Scientific data, vol. 5, no. 1, pp. 1–9, 2018

  61. [69]

    Skin lesion analysis toward melanoma detection 2018: A challenge hosted by the international skin imaging collaboration (isic),

    N. Codella, V . Rotemberg, P. Tschandl, M. E. Celebi, S. Dusza, D. Gutman, B. Helba, A. Kalloo, K. Liopyris, M. Marchetti et al., “Skin lesion analysis toward melanoma detection 2018: A challenge hosted by the international skin imaging collaboration (isic),” arXiv preprint ar...

  62. [70]

    Ph 2-a dermoscopic image database for research and benchmarking,

    T. Mendonc ¸a, P. M. Ferreira, J. S. Marques, A. R. Marcal, and J. Rozeira, “Ph 2-a dermoscopic image database for research and benchmarking,” in EMBC. IEEE, 2013, pp. 5437–5440

  63. [71]

    Sam 2: Segment anything in images and videos,

    N. Ravi, V . Gabeur, Y .-T. Hu, R. Hu, C. Ryali, T. Ma, H. Khedr, R. R¨adle, C. Rolland, L. Gustafson et al., “Sam 2: Segment anything in images and videos,” in ICLR, 2025

  64. [72]

    Sam-adapter: Adapting segment anything in underperformed scenes,

    T. Chen, L. Zhu, C. Deng, R. Cao, Y . Wang, S. Zhang, Z. Li, L. Sun, Y . Zang, and P. Mao, “Sam-adapter: Adapting segment anything in underperformed scenes,” in ICCV, 2023, pp. 3367–3375

  65. [73]

    Improving the generalization of segmentation foundation model under distribution shift via weakly supervised adaptation,

    H. Zhang, Y . Su, X. Xu, and K. Jia, “Improving the generalization of segmentation foundation model under distribution shift via weakly supervised adaptation,” in CVPR, 2024, pp. 23 385–23 395

  66. [74]

    Encoder- decoder with atrous separable convolution for semantic image segmen- tation,

    L.-C. Chen, Y . Zhu, G. Papandreou, F. Schroff, and H. Adam, “Encoder- decoder with atrous separable convolution for semantic image segmen- tation,” in ECCV, 2018, pp. 801–818

  67. [75]

    Transunet: Rethinking the u-net architecture design for medical image segmentation through the lens of transformers,

    J. Chen, J. Mei, X. Li, Y . Lu, Q. Yu, Q. Wei, X. Luo, Y . Xie, E. Adeli, Y . Wanget al., “Transunet: Rethinking the u-net architecture design for medical image segmentation through the lens of transformers,” MedIA, vol. 97, p. 103280, 2024

  68. [76]

    Swin transformer: Hierarchical vision transformer using shifted windows,

    Z. Liu, Y . Lin, Y . Cao, H. Hu, Y . Wei, Z. Zhang, S. Lin, and B. Guo, “Swin transformer: Hierarchical vision transformer using shifted windows,” in ICCV, 2021, pp. 10 012–10 022

  69. [77]

    Transformers are SSMs: Generalized models and efficient algorithms through structured state space duality,

    T. Dao and A. Gu, “Transformers are SSMs: Generalized models and efficient algorithms through structured state space duality,” in ICML, 2024

  70. [78]

    Famba-v: Fast vision mamba with cross-layer token fusion,

    H. Shen, Z. Wan, X. Wang, and M. Zhang, “Famba-v: Fast vision mamba with cross-layer token fusion,” in ECCV Workshop, 2024

  71. [79]

    Segment anything in medical images,

    J. Ma, Y . He, F. Li, L. Han, C. You, and B. Wang, “Segment anything in medical images,” Nature Communications, vol. 15, no. 1, p. 654, 2024

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.