Pith. sign in

REVIEW 4 major objections 5 minor 23 references

Segmentation of Coronary Artery Stenosis in X-ray Angiography using Mamba Models

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Mamba-based U-Net reaches 68.79% F1 for coronary stenosis segmentation, an 11.8-point jump over the previous best.

desk verdict A benchmark paper with a new top F1 on ARCADE stenosis, but the metric definition is missing, so the headline improvement may not be apples-to-apples. read the letter →

arxiv 2412.02568 v1 pith:IF5M7QBI submitted 2024-12-03 eess.IV cs.AIcs.CV

Ali Rostami , Fatemeh Fouladi , Hedieh Sajedi This is my paper · ORCID
classification eess.IVcs.AIcs.CV
keywords coronaryarterystenosissegmentationX-rayangiographyMambastatespacemodelsU-Netarchitectureselectivescan2DARCADEdatasetmedicalimageF1score
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

This paper asks whether Mamba-based sequence models can localize coronary artery stenosis—the narrowed, often blurry sections of arteries—in X-ray angiography images. It trains six U-shaped segmentation models on the ARCADE dataset and reports that U-Mamba BOT, a U-Net with Mamba blocks only in the bottleneck, reaches an F1 score of 68.79% on the 300-image test set. That is an 11.8-point improvement over the semi-supervised SSASS method, the strongest prior result the paper quotes. The practical stake is that a fast, linear-complexity model could support automated stenosis screening as a second reader for cardiologists.

What carries the argument

The central object is U-Mamba BOT, a U-shaped network that keeps convolution in the encoder and decoder and inserts Mamba blocks—selective structured state-space models with an input-dependent selection mechanism and scan-based computation—only at the bottleneck. The Mamba block processes image features in 1D sequences (via the 2D selective scan SS2D and cross-scan module from VMamba), giving linear scaling in sequence length rather than the quadratic attention cost of Transformers. This bottleneck placement is what the paper credits for combining global context with efficient training.

What would settle it

Re-run Ensemble Learning, StenUNet, and SSASS on the same 300-image ARCADE test images with identical preprocessing, metric computation, and thresholding; if any of them matches or exceeds 68.79% F1, or if U-Mamba BOT's score drops on a corrected split, the paper's central improvement claim fails. A secondary check is to evaluate U-Mamba BOT on an independent coronary angiography dataset with similar stenosis annotations.

Watch

Extended reading notes

Core claim

On the ARCADE stenosis segmentation benchmark, U-Mamba BOT—a U-Net encoder-decoder whose encoder and decoder use convolutional residual blocks while Mamba selective state-space blocks sit in the bottleneck—obtains the best F1 (68.79%) and recall (67.69%) among the tested models, with U-Mamba ENC posting the best precision (71.13%). The paper interprets this as evidence that Mamba's linear-complexity selective scan can capture the long-range context needed to find small, blurred stenotic lesions while avoiding the quadratic cost of transformers.

Load-bearing premise

The improvement over prior work rests on the assumption that the F1 scores quoted for Ensemble Learning (39.41%), StenUNet (53.48%), and SSASS (56.99%) were computed on the same 300-image ARCADE test split with the same evaluation protocol, because the paper did not re-run those methods.

Editorial extensions

If this is right

  • U-Mamba BOT reaches 68.79% F1 for coronary stenosis segmentation, outperforming the quoted semi-supervised SSASS score of 56.99% by 11.8 points.
  • Among Mamba variants, Swin-UMamba D shows that putting Mamba blocks on both encoder and decoder sides yields strong performance with fewer parameters (27M versus 60M for Swin-UMamba).
  • The linear-complexity Mamba family is presented as a practical alternative to vision transformers for imbalanced medical segmentation tasks where lesion regions are tiny.
  • The models are trained with 5-fold cross-validation on 1200 ARCADE images and tested on the standardized 300-image set, with plaques defined by SYNTAX Score criteria.

Reading between the lines

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

  • If the claimed comparison is valid, Mamba-based segmentation could be paired with post-processing or ensemble steps to push stenosis localization closer to clinical usability, since even the best F1 leaves room for false positives and false negatives.
  • The paper's success with bottleneck-only Mamba suggests that global context matters most at the coarsest resolution; a natural test is whether adding Mamba blocks to skip connections or decoder levels improves recall further.
  • A direct head-to-head re-run of Ensemble Learning, StenUNet, and SSASS under identical training and evaluation settings would settle whether the 11.8-point gain reflects architecture choice rather than differences in protocol.
  • Because the ARCADE dataset is public, the U-Mamba BOT result is reproducible and could serve as a baseline for future stenosis segmentation work, including external validation on other angiography datasets.
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 / 5 minor

Summary. The paper evaluates five Mamba-based U-Net variants (U-Mamba BOT, U-Mamba ENC, LightM-UNet, Swin-Umamba, Swin-Umamba D) and one Swin Transformer model (Swin UNetR) for coronary artery stenosis segmentation on the public ARCADE dataset. Using 1200 training images with 5-fold cross-validation and a held-out 300-image test set, the authors report that U-Mamba BOT achieves the best F1 score of 68.79%, an 11.8-point improvement over the previous SSASS result of 56.99%. The paper is an empirical benchmark comparison; it does not introduce a new architecture or release code.

Significance. If the comparison is valid, the paper provides useful evidence that Mamba-based U-Net backbones can improve stenosis localization in X-ray angiography on a standardized public benchmark. Strengths include the use of a public dataset, a held-out test set, and internal arithmetic consistency between the reported precision, recall, and F1 values in Table 1. The significance, however, depends critically on whether the F1 scores are computed with the same metric and evaluation protocol as the cited prior work; this is not documented in the manuscript. The apparent 11.8-point improvement over SSASS is therefore not yet established.

major comments (4)
  1. [Section 5, Table 1] The F1 metric is never defined. The manuscript must state whether F1 is computed per-pixel, per-connected-component, or per-lesion; give the binarization threshold and matching/overlap rule; and specify any post-processing. The cited prior results (39.41%, 53.48%, 56.99%) are benchmark numbers from the ARCADE challenge literature and were likely produced with the official ARCADE evaluation protocol. If the authors used a plain pixel-wise F1/Dice instead of that protocol, the numbers in Table 1 are not comparable and the claimed 11.8-point improvement over SSASS could be a metric artifact. This is a load-bearing point for the central state-of-the-art claim.
  2. [Section 5, Table 1] Only single-run results are reported, with no confidence intervals, standard deviations, or significance tests. U-Mamba BOT's F1 of 68.79% is only 0.22 percentage points above U-Mamba ENC (68.57%) and about 2.1 points above Swin-Umamba (66.82%). Without multiple runs or a statistical comparison, the ranking of the best model and the claimed improvement over second-best are not robust.
  3. [Sections 3 and 4] The experimental setup is described too sparsely for reproducibility. The paper does not report the loss function, optimizer, learning rate, batch size, number of epochs, data augmentation, or the rule for selecting the final model from 5-fold cross-validation. It also does not clarify whether the 300-image test set is exactly the official ARCADE test split or whether any of the 1200 training images overlap with validation. The authors should either release code/configuration files or provide these details so that the reported F1 values can be independently checked.
  4. [Table 1, Figure 4] The parameter counts should be verified. In particular, U-Mamba BOT is reported as 500M parameters while U-Mamba ENC is 104M; this is disproportionate and inconsistent with the public U-Mamba configurations, which typically place U-Mamba BOT far below 500M. Since Figure 4 uses parameter counts for the efficiency comparison, an erroneous count would alter the figure's message.
minor comments (5)
  1. [Section 5] The text says 'In this figure, we can see...' when referring to Table 1; this should be corrected.
  2. [Table 1 and Section 2] The model name is written inconsistently as 'Sten UNet', 'StenUNet', and 'StenUNet' in different places; standardize the spelling.
  3. [Section 2] The phrase 'an F1-score as high as 0.5348 on a test set only 0.0005 points less than that of the 2nd place' is ambiguous; it should specify which test set and which rank is being compared.
  4. [Section 4.2] There are grammatical errors such as 'we will going to discuss' and 'The encoder expands... four parts'; these should be corrected in a polished revision.
  5. [Figure 4] The caption's instruction to 'check the central point of each circle' is unclear; provide axis labels, a legend, and an explicit explanation of what the circle area represents.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an external benchmark evaluation whose central F1 result is a model output on a public dataset, not a quantity derived from or fitted to its inputs.

full rationale

The paper contains no derivation chain of the kind that would support a circularity finding. Its central claim is that U-Mamba BOT reaches an F1 score of 68.79% on the ARCADE stenosis segmentation test set, outperforming cited prior methods. This number is produced by training architecture variants taken from external prior work on the public ARCADE dataset, and the comparison baselines are cited from other research groups, not from the present authors. No parameter is fitted to the cited F1 values, and no equation defines the target result in terms of the inputs. The cited baselines (Ensemble Learning, StenUNet, SSASS) are not re-run by the authors, so the comparability of evaluation protocols is an open correctness and reproducibility concern, but it is not circularity: the paper does not use those baselines to construct its own predictions. There are also no load-bearing self-citations, since the authors cite no prior work of their own as justification for the method. The potential issue that the paper never defines its exact F1 computation or matching criterion is a metric-definition and verification problem, not a case where the claimed result reduces by construction to the input. Under the hard rules, circularity may be claimed only when the paper's own equations or self-citation chain force the result; no such reduction can be quoted here. The honest finding is therefore no significant circularity.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new theory, no new architectural components, and no fitted physical constants. Its central claim is an empirical performance number, so the ledger is minimal. The main unstated inputs are the correctness of the ARCADE annotations, the validity of the test split, and the comparability of prior published F1 scores. The missing training configuration is the largest 'free parameter' in the study.

free parameters (1)
  • Training hyperparameters and model selection protocol = Not reported
    The paper does not specify learning rate, batch size, number of epochs, loss function, optimizer, data augmentation, or how the final model was chosen from the 5 cross-validation folds. These choices directly influence the reported F1 scores and are effectively free parameters of the experiment.
assumptions (3)
  • domain assumption ARCADE dataset annotations are accurate and follow the SYNTAX Score definition (stenosis with thickness > 1.5 mm and >= 50% narrowing is labeled as a lesion).
    Section 3 states the labels were produced by medical professionals using this definition; the paper relies on these labels as ground truth.
  • domain assumption The official ARCADE test split of 300 images is used for final evaluation and was not used for training or validation.
    Section 3 says 'the rest 300 for testing the final model'; no leakage prevention is described beyond this statement.
  • domain assumption Reported F1 scores of prior methods (Ensemble Learning, StenUNet, SSASS) are directly comparable to this paper's results because they use the same dataset and evaluation metric.
    Table 1 compares to these numbers without re-running those methods, so the comparison assumes identical test splits and protocols.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Segmentation of Coronary Artery Stenosis in X-ray Angiography using Mamba Models." pith.science (2026). https://pith.science/paper/IF5M7QBI

@misc{pith2026241202568,
  author       = {Pith},
  title        = {Pith review of: Segmentation of Coronary Artery Stenosis in X-ray Angiography using Mamba Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IF5M7QBI}},
  note         = {Machine review of arXiv:2412.02568}
}
read the original abstract

Coronary artery disease stands as one of the primary contributors to global mortality rates. The automated identification of coronary artery stenosis from X-ray images plays a critical role in the diagnostic process for coronary heart disease. This task is challenging due to the complex structure of coronary arteries, intrinsic noise in X-ray images, and the fact that stenotic coronary arteries appear narrow and blurred in X-ray angiographies. This study employs five different variants of the Mamba-based model and one variant of the Swin Transformer-based model, primarily based on the U-Net architecture, for the localization of stenosis in Coronary artery disease. Our best results showed an F1 score of 68.79% for the U-Mamba BOT model, representing an 11.8% improvement over the semi-supervised approach.

Figures

Figures reproduced from arXiv: 2412.02568 by the authors.

Figure 1
Figure 1. The Mamba architecture [6] While the methodology that Mamba follows is closest to the recurrent models, what differentiates the approach from others is in terms of the selection mechanism in [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Methodology used in this study. The input images and corresponding masks are fed into various model architectures, [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Visualization of each model’s performance on the ARCADE dataset [20]. The Y-axis of this chart shows their per￾formance you have to check the central point of each circle for it and circle diameters represent the number of parameters used in each of these models. 5. Results In this section, we will evaluate the models against certain metrics. The most important metrics for us here were precision and recall so let’s … view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Prediction of different models on test dataset. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 12 canonical work pages

  1. [1]

    Automatic detection and classification of cardiovascular disorders using phono- cardiogram and convolutional vision transformers

    Qaisar Abbas, Ayyaz Hussain, and Abdul Rauf Baig. Automatic detection and classification of cardiovascular disorders using phono- cardiogram and convolutional vision transformers. diagnostics, 12(12):3109, 2022

    work page 2022

  2. [2]

    Multivessel coronary artery seg- mentation and stenosis localisation using ensemble learning

    Muhammad Bilal, Dinis Martinho, Reiner Sim, Adnan Qayyum, Hunaid V ohra, Massimo Caputo, Taofeek Akinosho, Sofiat Abioye, Zaheer Khan, Waleed Niaz, et al. Multivessel coronary artery seg- mentation and stenosis localisation using ensemble learning. arXiv preprint arXiv:2310.17954, 2023

    arXiv

  3. [3]

    Swin-unet: Unet-like pure transformer for medical image segmentation

    Hu Cao, Yueyue Wang, Joy Chen, Dongsheng Jiang, Xiaopeng Zhang, Qi Tian, and Manning Wang. Swin-unet: Unet-like pure transformer for medical image segmentation. In European conference on computer vision, pages 205–218. Springer, 2022

    2022

  4. [4]

    Automated stenosis detection and classification in x-ray angiography using deep neural network

    Chao Cong, Yoko Kato, Henrique Doria Vasconcellos, Joao Lima, and Bharath Venkatesh. Automated stenosis detection and classification in x-ray angiography using deep neural network. In 2019 IEEE international conference on bioinformatics and biomedicine (BIBM) , pages 1301–1308. IEEE, 2019

    work page 2019

  5. [5]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weis- senborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at scale. 2021

    work page 2021

  6. [6]

    Mamba: Linear-time sequence modeling with selective state spaces, 2024

    Albert Gu and Tri Dao. Mamba: Linear-time sequence modeling with selective state spaces, 2024

    2024

  7. [7]

    Efficiently modeling long sequences with structured state spaces, 2021

    Albert Gu, Karan Goel, and Christopher Ré. Efficiently modeling long sequences with structured state spaces, 2021

    work page 2021

  8. [8]

    Swin unetr: Swin transformers for semantic segmentation of brain tumors in mri images

    Ali Hatamizadeh, Vishwesh Nath, Yucheng Tang, Dong Yang, Hol- ger R Roth, and Daguang Xu. Swin unetr: Swin transformers for semantic segmentation of brain tumors in mri images. In International MICCAI brainlesion workshop , pages 272–284. Springer, 2021

    work page 2021

Show all 23 references
  1. [9]

    Vision transformer in stenosis detection of coronary arteries

    Michał Jungiewicz, Piotr Jastrz˛ ebski, Piotr Wawryka, Karol Przys- talski, Karol Sabatowski, and Stanisław Bartu ´s. Vision transformer in stenosis detection of coronary arteries. Expert Systems with Applications, 228:120234, 2023

    2023

  2. [10]

    Ssass: Semi-supervised approach for stenosis segmentation

    In Kyu Lee, Junsup Shin, Yong-Hee Lee, Jonghoe Ku, and Hyun-Woo Kim. Ssass: Semi-supervised approach for stenosis segmentation. arXiv preprint arXiv:2311.10281 , 2023

    2023 arXiv

  3. [11]

    Lightm-unet: Mamba assists in lightweight unet for medical image segmentation

    Weibin Liao, Yinghao Zhu, Xinyuan Wang, Chengwei Pan, Yasha Wang, and Liantao Ma. Lightm-unet: Mamba assists in lightweight unet for medical image segmentation. arXiv preprint arXiv:2403.05246, 2024

    2024 arXiv

  4. [12]

    Pathophysiology of coronary artery disease

    Peter Libby and Pierre Theroux. Pathophysiology of coronary artery disease. Circulation, 111(25):3481–3488, 2005

    2005

  5. [13]

    Ste- nunet: automatic stenosis detection from x-ray coronary angiography

    Hui Lin, Tom Liu, Aggelos Katsaggelos, and Adrienne Kline. Ste- nunet: automatic stenosis detection from x-ray coronary angiography. arXiv preprint arXiv:2310.14961 , 2023

    2023 arXiv

  6. [14]

    Swin- umamba: Mamba-based unet with imagenet-based pretraining

    Jiarun Liu, Hao Yang, Hong-Yu Zhou, Yan Xi, Lequan Yu, Cheng Li, Yong Liang, Guangming Shi, Yizhou Yu, Shaoting Zhang, et al. Swin- umamba: Mamba-based unet with imagenet-based pretraining. In In- ternational Conference on Medical Image Computing and Computer- Assisted Interve...

    2024

  7. [15]

    Vmamba: Visual state space model, 2024

    Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, and Yunfan Liu. Vmamba: Visual state space model, 2024

    2024

  8. [16]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF international conference on computer vision , pages 10012–10022, 2021

    2021

  9. [17]

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

    Jun Ma, Feifei Li, and Bo Wang. U-mamba: Enhancing long- range dependency for biomedical image segmentation. arXiv preprint arXiv:2401.04722, 2024

    2024 arXiv

  10. [18]

    An introduction to convolutional neural networks

    K O’Shea. An introduction to convolutional neural networks. arXiv preprint arXiv:1511.08458, 2015

    2015 arXiv

  11. [19]

    Hybrid classical–quantum con- volutional neural network for stenosis detection in x-ray coronary angiography

    Emmanuel Ovalle-Magallanes, Juan Gabriel Avina-Cervantes, Ivan Cruz-Aceves, and Jose Ruiz-Pinales. Hybrid classical–quantum con- volutional neural network for stenosis detection in x-ray coronary angiography. Expert Systems with Applications , 189:116112, 2022

    2022

  12. [20]

    Dataset for automatic region-based coronary artery disease diagnos- tics using x-ray angiography images

    Maxim Popov, Akmaral Amanturdieva, Nuren Zhaksylyk, Alsabir Alkanov, Adilbek Saniyazbekov, Temirgali Aimyshev, Eldar Ismailov, Ablay Bulegenov, Arystan Kuzhukeyev, Aizhan Kulanbayeva, et al. Dataset for automatic region-based coronary artery disease diagnos- tics using x-ray a...

    2024

  13. [21]

    U-net: Con- volutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Con- volutional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, par...

    2015

  14. [22]

    Atten- tion is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Atten- tion is all you need. advances in neural information processing sys- tems. Advances in neural information processing systems , 30(2017), 2017

    2017

  15. [23]

    Automated identification and grading of coronary artery stenoses with x-ray angiography

    Tao Wan, Hongxiang Feng, Chao Tong, Deyu Li, and Zengchang Qin. Automated identification and grading of coronary artery stenoses with x-ray angiography. Computer methods and programs in biomedicine , 167:13–22, 2018

    2018

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.