Pith. sign in

REVIEW 3 major objections 5 minor 39 references

BATseg: Boundary-aware Multiclass Spinal Cord Tumor Segmentation on 3D MRI Scans

T0 review · 3 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 Adding per-class surface distance maps to a segmentation backbone improves spinal cord tumor segmentation, backed by a new 653-patient MRI dataset

desk verdict The spinal cord tumor dataset is a genuine contribution, but the published boundary-aware loss has an inverted sign that makes the method as written impossible; needs major revision. read the letter →

arxiv 2412.06507 v1 pith:JCS3JC6D submitted 2024-12-09 eess.IV cs.CVcs.LG

Hongkang Song , Zihui Zhang , Yanpeng Zhou , Jie Hu , Zishuo Wang , Hou Him Chan , Chon Lok Lei , Chen Xu
show 2 more authors
Yu Xin Bo Yang
This is my paper · ORCID
classification eess.IVcs.CVcs.LG
keywords spinalcordtumorsegmentation3DMRIboundary-awarelosssurfacedistancefieldmulticlassnnUNetmedicalimageKNIGHTkidneydataset
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 tries to establish that the hard part of segmenting spinal cord tumors is their boundaries, and that a network can learn those boundaries by regressing a per-class distance-to-surface map rather than only labeling voxels. To test this, the authors built the first large-scale multiclass spinal cord tumor segmentation dataset, 653 patients with gadolinium-enhanced T1-weighted 3D MRI covering meningiomas, ependymomas, astrocytomas, and hemangioblastomas, split into five folds. They augmented the nnUNet backbone with a second head that predicts a truncated, normalized, class-wise tumor surface distance field, supervised by a boundary-aware loss in addition to cross-entropy and Dice. They report that this augmented model beats the backbone and several transformer and two-stage baselines on both their dataset and the public KNIGHT kidney tumor dataset, with mean Dice on the spinal dataset rising from 52.8% to 62.1% and Hausdorff distance falling from 161.4 mm to 108.3 mm. If correct, the result gives clinicians a more accurate automatic way to measure tumor size, location, and type, and gives the field a reusable geometric prior for other small, boundary-dominated lesions.

What carries the argument

The load-bearing object is the class-wise tumor surface distance field: a per-voxel map, computed from the ground-truth mask for each tumor type separately, in which voxels inside the tumor store their positive nearest distance to the tumor boundary, voxels outside store negative distances truncated to zero beyond the maximum interior distance, and all retained values are normalized into [0,1]. It is supervised by the multiclass boundary-aware loss, which is the negative of the squared prediction error times the absolute prediction error, averaged over the volume; this focuses optimization on the nonzero boundary band instead of the large zero background. The field is predicted by a single convolutional head running in parallel with the segmentation head of the nnUNet backbone, and the three losses (cross-entropy, Dice, and boundary-aware) are trained jointly with equal weights. The distance field is the mechanism that makes the segmentation head's predicted surfaces contract toward the annotated tumor boundary.

What would settle it

Train the same nnUNet backbone with an auxiliary head that regresses a scrambled version of the distance field, or an unrelated per-voxel target, under the same loss weighting; if Dice and Hausdorff gains match BATseg's, the boundary-distance geometry is not the operative cause. A second test is to re-run the five-fold evaluation on masks re-annotated by a different center or scanner: if the truncation and normalization ablations stop favoring the full recipe, the claim that the distance field carries the gain is falsified.

Watch

Extended reading notes

Core claim

On its own terms, the paper's claim is that multiclass tumor segmentation is improved when the network is forced, in parallel with ordinary voxel labeling, to reconstruct a surface distance field for each tumor class separately. The field is defined per class from the manual mask: voxels inside the tumor carry their positive nearest distance to the boundary, voxels outside carry negative distances until truncation at the maximum interior distance, and surviving values are normalized to [0,1]. A head on the nnUNet backbone regresses this field under a loss that weights voxels by the product of squared and absolute error, so the optimization concentrates on nonzero (near-boundary) voxels; the authors report this branch is what tightens predicted surfaces. The evidence is experimental: across five folds of the new spinal cord dataset, BATseg reports mean Dice 62.1% versus 52.8% for the nnUNet backbone and mean Hausdorff distance 108.3 mm versus 161.4 mm, and on the KNIGHT kidney dataset it reports mean Dice 40.7% versus 33.8% for nnUNet. Ablations attribute the gains to truncation, normalization, the multiclass (class-wise) definition of the field, and the squared-error weighting in the boundary-aware loss.

Load-bearing premise

The reported improvement rests on the assumption that the manual tumor outlines are precise enough that forcing the network to also learn a per-type map of distance-to-boundary adds real geometric information beyond what ordinary voxel-labeling losses already provide.

Editorial extensions

If this is right

  • On the new 653-patient spinal cord tumor dataset, BATseg reports a mean Dice of 62.1% versus 52.8% for its nnUNet backbone, and a mean Hausdorff distance of 108.3 mm versus 161.4 mm across the four tumor types.
  • On the public KNIGHT kidney tumor dataset, the same recipe reports mean Dice 40.7% versus 33.8% for nnUNet, indicating the boundary-distance loss transfers to a different tumor domain.
  • Per-class distance fields outperform a class-agnostic field on KNIGHT by 4.3 points in mean Dice, supporting the design choice to treat each tumor type's surface separately.
  • Ablations show that truncating the outside-distance field, normalizing the retained values, and weighting the loss by squared error each contribute to the reported gains.
  • Segmentation remains hardest for astrocytomas (35.2% Dice), which the paper attributes to their similar appearance to ependymomas and resulting misclassification.

Reading between the lines

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

  • Editorial inference: the class-wise distance-field recipe is not tested on other small-lesion segmentation tasks, but the KNIGHT transfer result suggests it is a candidate drop-in auxiliary loss for any multiclass segmentation where boundaries rather than bulk volume carry the diagnostic signal.
  • Editorial inference: because the paper does not include a control branch with an unrelated auxiliary regression target, the reported gains are not yet isolated to the geometric content of the distance field; a scrambled-distance control would settle that.
  • Editorial inference: a natural next experiment is re-annotation or cross-scanner evaluation; the paper explicitly lists scanner generalization as untested, so a held-out scanner study would show whether the distance-field prior survives realistic clinical variation.
  • Editorial inference: the dataset's class imbalance and small tumor volumes suggest that reporting per-class precision-recall or volume error, beyond Dice and Hausdorff, would clarify where the boundary head helps most.
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

3 major / 5 minor

Summary. The paper proposes BATseg, a method that augments an nnUNet backbone with an auxiliary head that predicts a truncated, normalized, class-wise tumor surface distance field, supervised by a new "boundary-aware" loss. The authors also introduce a new dataset of 653 gadolinium-enhanced T1-weighted 3D MRI scans of spinal cord tumors with four tumor types and five-fold splits. On this dataset and on the public KNIGHT kidney-tumor dataset, BATseg is reported to outperform several end-to-end and two-stage baselines in Dice score and 95th-percentile Hausdorff distance. The paper includes per-fold results, ablations on KNIGHT, and qualitative visualizations.

Significance. If the results hold, the new spinal cord tumor dataset and the simple auxiliary-surface-distance training strategy would be useful contributions to medical image segmentation, particularly for small, irregular tumor structures. The paper provides consistent five-fold cross-validation results and ablation studies on a public dataset, which are strengths. However, the central method as written contains two internal inconsistencies in the definition of the ground-truth distance field and the boundary-aware loss, the most serious being a sign error in Eq. (3) that makes the printed loss non-positive and repulsive from the target. As a result, the reported experiments cannot be reproduced from the described objective, and the attribution of the observed improvements to the proposed loss is not supported as written. These issues are correctable, but they are load-bearing for the manuscript's central claim.

major comments (3)
  1. [Sec. 4.3, Eq. (3)] The boundary-aware loss as printed, ℓba = -(f - \bar f)^2 |f - \bar f|, is non-positive and unbounded below. Its gradient with respect to f is -3|f-\bar f|(f-\bar f), so gradient descent pushes f away from \bar f: the loss is maximized, not minimized, when f = \bar f. This cannot be the objective used in the experiments, because Table 8 shows that adding the squared term improves Dice by about 1%, whereas the printed negative sign would make the squared term actively harmful. Please correct the sign in Eq. (3) and explicitly confirm that the reported results were obtained with the corrected objective.
  2. [Sec. 4.2, Eq. (2)] The text states that "all previously truncated values" are kept as zeros, but the normalization formula in Eq. (2) does not do this: for a truncated voxel with d'_j = 0, the equation gives \bar d_j = (0 + 1)/2 = 0.5. Thus truncated background voxels are labeled 0.5, not 0, which contradicts the description in Sec. 4.3 that background voxels "tend to be easily optimized to be zeros" and shifts the target distribution of the distance-field head. Please correct Eq. (2) or the accompanying description, and specify the actual normalization used in the experiments.
  3. [Sec. 5.3, Tables 5-8] The ablation studies vary properties of the distance field (truncation, normalization, class-wise vs. class-agnostic, loss form) but do not include a control branch with an auxiliary target that carries no geometric information (e.g., a constant, random, or unrelated head). Without such a control, the improvement over nnUNet could be attributed to the extra parameters or to the presence of an auxiliary supervision signal in general, rather than specifically to the truncated normalized tumor-surface distance field. Adding this control would substantially strengthen the central causal claim.
minor comments (5)
  1. [Sec. 5.3] The word "emperically" should be "empirically".
  2. [Author affiliations] In the author block, "Insititute" and "Acadamy" should be "Institute" and "Academy".
  3. [Data availability] The GitHub link in Sec. 1 is written as "https://github.com/vLAR-group/BA T seg" with a space; please provide a working, correctly spaced URL.
  4. [Tables 3-8] The tables list "Hausdorff Distance (HD)" but the metrics paragraph in Sec. 5 specifies the 95th-percentile Hausdorff distance; please make the captions consistent with the metric definition.
  5. [Table 4] The rows for nnUNet+UEnc with mask and nnUNet+UEnc w/o mask are identical (NoAT Dice 50.5, CanAT Dice 0.0); please explain why they coincide or correct the results, since the text describes them as distinct methods.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: BATseg's auxiliary distance-field supervision is a standard multi-task target; the central concern is an apparent sign error in Eq. 3, which is a correctness issue rather than circularity.

full rationale

I find no step in the paper where a reported prediction is equivalent to an input by construction. The method is a standard supervised training setup: the segmentation head is trained with cross-entropy and Dice losses against manual labels, and the auxiliary surface-distance head is trained against a deterministic, class-wise truncated and normalized distance transform of the same manual labels. This is a legitimate auxiliary supervision target: the network must still infer the distance field from the input MRI volume, and all reported results are computed on held-out folds (spinal cord dataset) or the fixed KNIGHT test split (KNIGHT dataset). The fact that the distance-field label is a re-encoding of the ground-truth mask does not make the evaluation circular, because the auxiliary label is not the evaluation metric and the segmentation output is not derived from the distance-field head at inference. No fitted parameter is renamed as a prediction, no self-cited uniqueness theorem is invoked, and no ansatz is smuggled in via citation. I also considered the KNIGHT ablation protocol: Tables 5-8 are performed on the KNIGHT dataset, and if those ablations were used to select the final design on the KNIGHT test set, that would be an evaluation-protocol concern, but it is not an equation-level circularity under the rubric. Finally, I flag the apparent sign error in Eq. 3 as a serious correctness and reproducibility problem that is outside the circularity definition: as printed, Eq. 3 simplifies to -|f - f_bar|^3, whose gradient descent moves predictions away from the ground-truth distance values, directly contradicting the paper's statement that the boundary-aware loss drives the branch to learn the distance field. This should be corrected and re-verified, but it does not constitute a circular derivation, so the circularity score remains 0.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The method rests on standard supervised-learning assumptions and the reliability of the manual annotations. There are no fitted constants that the central claim depends on beyond the neural network weights themselves; the hand-chosen hyperparameters (truncation multiplier, equal loss weights) are ablated on a separate public dataset.

free parameters (2)
  • Truncation multiplier = 1 (ablated: 2, 3)
    The truncation threshold is set to one times the maximum inside distance; the authors ablate this choice on KNIGHT (Table 6) but provide no principled criterion.
  • Loss weights = 1:1:1 (CE, Dice, boundary-aware)
    Equal weights are chosen 'for simplicity' in Eq. 4; no tuning or sensitivity analysis is reported.
assumptions (3)
  • domain assumption Manual tumor annotations are accurate and consistent across the four tumor types.
    The distance-field targets derive entirely from these annotations; the paper notes astrocytoma/ependymoma misclassification is the main source of low Dice for those classes (Section 5.1).
  • domain assumption Resampled 3D volumes preserve the tumor surface geometry despite original slice thicknesses of up to 8 mm.
    Preprocessing (Section 3) interpolates intensities and masks; for small tumors, the slice-direction distance field is computed from interpolated labels and could be noisy.
  • standard math The distance transform and its truncation/normalization produce a well-posed training target.
    These are standard operations, but the printed normalization equation is garbled, so the exact target is ambiguous as published.

how reviews work

0 comments
Cite this review

Pith. "Pith review of BATseg: Boundary-aware Multiclass Spinal Cord Tumor Segmentation on 3D MRI Scans." pith.science (2026). https://pith.science/paper/JCS3JC6D

@misc{pith2026241206507,
  author       = {Pith},
  title        = {Pith review of: BATseg: Boundary-aware Multiclass Spinal Cord Tumor Segmentation on 3D MRI Scans},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JCS3JC6D}},
  note         = {Machine review of arXiv:2412.06507}
}
read the original abstract

Spinal cord tumors significantly contribute to neurological morbidity and mortality. Precise morphometric quantification, encompassing the size, location, and type of such tumors, holds promise for optimizing treatment planning strategies. Although recent methods have demonstrated excellent performance in medical image segmentation, they primarily focus on discerning shapes with relatively large morphology such as brain tumors, ignoring the challenging problem of identifying spinal cord tumors which tend to have tiny sizes, diverse locations, and shapes. To tackle this hard problem of multiclass spinal cord tumor segmentation, we propose a new method, called BATseg, to learn a tumor surface distance field by applying our new multiclass boundary-aware loss function. To verify the effectiveness of our approach, we also introduce the first and large-scale spinal cord tumor dataset. It comprises gadolinium-enhanced T1-weighted 3D MRI scans from 653 patients and contains the four most common spinal cord tumor types: astrocytomas, ependymomas, hemangioblastomas, and spinal meningiomas. Extensive experiments on our dataset and another public kidney tumor segmentation dataset show that our proposed method achieves superior performance for multiclass tumor segmentation.

Figures

Figures reproduced from arXiv: 2412.06507 by the authors.

Figure 1
Figure 1. An illustration of four types of spinal cord tumors shown in the first row, and other commonly studied organs/tumors shown in the second row. excellent segmentation performance on a variety of medical images, thanks to the availability of large-scale datasets [23, 28]. However, there is still a lack of an automatic model that can precisely segment and recognize multiple types of spinal cord tumors. The main reasons … view at source ↗
Figure 2
Figure 2. An illustration of the overall framework. ticlass boundary-aware loss function that aids the backbone network such as nnUNet [14] to precisely segment multiple types of spinal cord tumors. In particular, as illustrated in [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The proposed segmentation pipeline. A 3D volume T is fed into the backbone network, predicting per-voxel multiclass results S via the segmentation head, and es￾timating the tumor surface distance field F via a newly added head. d𝒊 𝒑𝒊 𝒑𝒋 d𝒋 MRI Slice t Tumor Mask s Distance Calculation Normalized Truncated Distance Field 0 1 [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: An illustration of calculating tumor surface distance values on a single 2D slice. 4 Method 4.1 Overview As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: The two-stage baseline for our problem of multiclass 3D segmentation. Given an input 3D volume, class-agnostic tumor voxels are segmented against the background in Stage 1, followed by Stage 2 where the tumor type is classified for the same 3D volume. The predicted tum…
Figure 6
Figure 6. Figure 6: Qualitative results of different ablation studies on the public KNIGHT dataset. Group 1/2/3/4 ablations are in the red, blue, green, and pink dotted boxes respectively. – We notice that some two-stage methods (nnUNet+UEncwith mask and nnUNet+ UEncw/o mask) tend to clas…
Figure 7
Figure 7. Figure 7: Qualitative comparison of all methods on our spinal cord tumor dataset and the KNIGHT dataset [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 35 canonical work pages

  1. [1]

    MICCAI (2018)

    B, S.M., Mercan, E., Bartlett, J., Weaver, D.: Y-Net: Joint Segmentation and Classification for Diagnosis. MICCAI (2018)

    work page 2018

  2. [2]

    SSRN (2023)

    Barros, V., Abdallah, N., Raboh, M., Heller, N., Rabinovici-Cohen, S., Golts, A., Weight, C.: Preoperative Kidney Tumor Risk Estimation from CT Imaging for Adjuvant Treatment Prediction. SSRN (2023)

    work page 2023

  3. [3]

    FlowSDF: Flow Matching for Medical Image Segmentation Using Distance Transforms

    Bogensperger, L., Narnhofer, D., Falk, A., Schindler, K., Pock, T.: FlowSDF: Flow Matching for Medical Image Segmentation Using Distance Transforms. arXiv:2405.18087 (2024)

    work page Pith review arXiv 2024

  4. [4]

    CVPR (2021)

    Borse, S., Wang, Y., Zhang, Y., Porikli, F.: InverseForm: A Loss Function for Structured Boundary-Aware Segmentation. CVPR (2021)

    work page 2021

  5. [5]

    IEEE Transactions on Medical Imaging (2021)

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

    work page 2021

  6. [6]

    arXiv:2102.04306 (2021)

    Chen,J.,Lu,Y.,Yu,Q.,Luo,X.,Adeli,E.,Wang,Y.,Lu,L.,Yuille,A.L.,Zhou,Y.: TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation. arXiv:2102.04306 (2021)

    arXiv 2021

  7. [7]

    MICCAI (2016)

    Cicek, O., Abdulkadir, A., Lienkamp, S.S., Brox, T., Ronneberger, O.: 3D U-Net: Learning Dense Volumetric. MICCAI (2016)

    work page 2016

  8. [8]

    Dice,L.R.:Measuresoftheamountofecologicassociationbetweenspecies.Ecology (1945)

    work page 1945

Show all 39 references
  1. [9]

    https://grand-challenge.org/ (2024)

    Grand Challenge, M.: A platform for end-to-end development of machine learning solutions in biomedical imaging. https://grand-challenge.org/ (2024)

    2024

  2. [10]

    MICCAI Workshop (2022)

    Hatamizadeh, A., Nath, V., Tang, Y., Yang, D., Roth, H.R., Xu, D.: Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images. MICCAI Workshop (2022)

    2022

  3. [11]

    WACV (2022)

    Hatamizadeh, A., Tang, Y., Nath, V., Yang, D., Myronenko, A., Landman, B., Roth, H.R., Xu, D.: UNETR: Transformers for 3D Medical Image Segmentation. WACV (2022)

    2022

  4. [12]

    CVPR (2016)

    He, K., Zhang, X., Ren, S., Sun, J.: Deep Residual Learning for Image Recognition. CVPR (2016)

    2016

  5. [13]

    ICASSP (2020) 16 H

    Huang,H.,Lin,L.,Tong,R.,Hu,H.,Zhang,Q.,Iwamoto,Y.,Han,X.,Chen,Y.W., Wu, J.: UNet 3+: A Full-Scale Connected UNet for Medical Image Segmentation. ICASSP (2020) 16 H. Song et al

    2020

  6. [14]

    Nature Methods (2020)

    Isensee, F., Jaeger, P.F., Kohl, S.A.A., Petersen, J., Maier-hein, K.H.: nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nature Methods (2020)

    2020

  7. [15]

    MIDL (2024)

    Islam, M.M., De Vente, C., Liefers, B., Klaver, C., Bekkers, E.J., Sánchez, C.I.: Uncertainty-aware retinal layer segmentation in OCT through probabilistic signed distance functions. MIDL (2024)

    2024

  8. [16]

    ICCV (2023)

    Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A.C., Lo, W.Y., Dollár, P., Girshick, R.: Segment Anything. ICCV (2023)

    2023

  9. [17]

    Springer (2018)

    Kirshblum, S., Lin, V.W.: Spinal cord medicine. Springer (2018)

    2018

  10. [18]

    ICLR (2023)

    Lee, H.H., Bao, S., Huo, Y., Landman, B.A.: 3D UX-Net: A Large Kernel Volu- metric ConvNet Modernizing Hierarchical Transformer for Medical Image Segmen- tation. ICLR (2023)

    2023

  11. [19]

    NeuroImage: Clinical (2021)

    Lemay, A., Gros, C., Zhuo, Z., Zhang, J., Duan, Y., Cohen-Adad, J., Liu, Y.: Automatic multiclass intramedullary spinal cord tumor segmentation on mri with deep learning. NeuroImage: Clinical (2021)

    2021

  12. [20]

    ICME (2019)

    Lin, F., Liu, C., Xie, H., Zha, Z.J., Zhang, Y.: Semantic-embedding and shape- aware U-net for ultrasound eyeball segmentation. ICME (2019)

    2019

  13. [21]

    ICCV (2023)

    Liu, J., Zhang, Y., Chen, J.N., Xiao, J., Lu, Y., Landman, B.A., Yuan, Y., Yuille, A., Tang, Y., Zhou, Z.: CLIP-Driven Universal Model for Organ Segmentation and Tumor Detection. ICCV (2023)

    2023

  14. [22]

    Nature Communications (2024)

    Ma, J., Wang, B.: Segment Anything in Medical Images. Nature Communications (2024)

    2024

  15. [23]

    TMI (2015)

    Menze, B.H., Jakab, A., Bauer, S.: The Multimodal Brain Tumor Image Segmen- tation Benchmark (BRATS). TMI (2015)

    2015

  16. [24]

    arXiv:2101.04704 (2021)

    Qin, X., Fan, D.P., Huang, C., Diagne, C., Zhang, Z., Sant’Anna, A.C., Suàrez, A., Jagersand, M., Shao, L.: Boundary-Aware Segmentation Network for Mobile and Web Applications. arXiv:2101.04704 (2021)

    2021 arXiv

  17. [25]

    In: Medi- calImaging2019:BiomedicalApplicationsinMolecular,Structural,andFunctional Imaging

    Reza, S.M., Roy, S., Park, D.M., Pham, D.L., Butman, J.A.: Cascaded convolu- tional neural networks for spine chordoma tumor segmentation from mri. In: Medi- calImaging2019:BiomedicalApplicationsinMolecular,Structural,andFunctional Imaging. SPIE (2019)

    2019

  18. [26]

    Springer Science & Business Media (2009)

    Rockafellar, R.T., Wets, R.J.B.: Variational analysis. Springer Science & Business Media (2009)

    2009

  19. [27]

    MICCAI (2015)

    Ronneberger, O., Fischer, P., Brox, T.: U-Net: Convolutional Networks for Biomed- ical Image Segmentation. MICCAI (2015)

    2015

  20. [28]

    arXiv:1902.09063 (2019)

    Simpson, A.L., Antonelli, M., Bakas, S.: A large annotated medical image dataset for the development and evaluation of segmentation algorithms. arXiv:1902.09063 (2019)

    2019 arXiv

  21. [29]

    Biologiske skrifter (1948)

    Sorensen, T.: A method of establishing groups of equal amplitude in plant soci- ology based on similarity of species content and its application to analyses of the vegetation on danish commons. Biologiske skrifter (1948)

    1948

  22. [30]

    MICCAI (2020)

    Valanarasu,J.M.J., Sindagi, V.A.,Hacihaliloglu, I., Patel, V.M.:KiU-Net: Towards Accurate Segmentation of Biomedical Images Using Over-Complete Representa- tions. MICCAI (2020)

    2020

  23. [31]

    BMVC (2021)

    V˜ u Ngoc, M.Ô., Chen, Y., Boutry, N., Chazalon, J., Carlinet, E., Fabrizio, J., Mallet, C., Géraud, T.: Introducing the Boundary-Aware loss for deep image seg- mentation. BMVC (2021)

    2021

  24. [32]

    JBHI (2021) BATseg: Boundary-aware Multiclass Spinal Cord Tumor Segmentation

    Wang, B., Wei, W., Qiu, S., Wang, S., Li, D., He, H.: Boundary Aware U-Net for Retinal Layers Segmentation in Optical Coherence Tomography Images. JBHI (2021) BATseg: Boundary-aware Multiclass Spinal Cord Tumor Segmentation ... 17

    2021

  25. [33]

    IET Image Processing (2022)

    Wang, R., Lei, T., Cui, R., Zhang, B., Meng, H., Nandi, A.K.: Medical image segmentation using deep learning: A survey. IET Image Processing (2022)

    2022

  26. [34]

    ECCV (2024)

    Wong, H.E., Rakic, M., Guttag, J., Dalca, A.V.: ScribblePrompt: Fast and Flexible Interactive Segmentation for Any Biomedical Image. ECCV (2024)

    2024

  27. [35]

    arXiv:2304.12620 (2023)

    Wu,J.,Fu,R.,Fang,H.,Liu,Y.,Wang,Z.,Xu,Y.,Jin,Y.,Arbel,T.:MedicalSAM Adapter: Adapting Segment Anything Model for Medical Image Segmentation. arXiv:2304.12620 (2023)

    2023 arXiv

  28. [36]

    Medical Image Analysis (2024)

    Zhang, M., Zhang, H., You, X., Yang, G.Z., Gu, Y.: Implicit Shape Modeling for Anatomical Structure Refinement of Volumetric Medical Images. Medical Image Analysis (2024)

    2024

  29. [37]

    ICCV (2019)

    Zhao, Y., Li, J., Zhang, Y., Tian, Y.: Multi-class part parsing with joint boundary- semantic awareness. ICCV (2019)

    2019

  30. [38]

    TIP (2023)

    Zhou, H.Y., Guo, J., Zhang, Y., Han, X., Yu, L., Wang, L., Yu, Y.: nnFormer: Volumetric Medical Image Segmentation via a 3D Transformer. TIP (2023)

    2023

  31. [39]

    TMI (2020) 18 H

    Zhou, Z., Siddiquee, M.M.R., Tajbakhsh, N., Liang, J.: UNet++: Redesigning Skip Connections to Exploit Multiscale Features in Image Segmentation. TMI (2020) 18 H. Song et al. 7 Appendix T able 9:The Dice Coefficient (%) and Hausdorff Distance (HD) (mm) scores on Fold 1 of our ...

    2020

Pith tools

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