UNetVL: Enhancing 3D Medical Image Segmentation with Chebyshev KAN Powered Vision-LSTM

Anant Madabhushi; Gourav Modanwal; Rohan Dhamdhere; Tanmoy Dam; Xuhui Guo

REVIEW 5 major objections 6 minor 22 references

UNetVL: Enhancing 3D Medical Image Segmentation with Chebyshev KAN Powered Vision-LSTM

T0 review · 5 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper claims that UNETVL, a U-Net with a Vision-LSTM encoder and Chebyshev-KAN projections, raises mean Dice by 7.3% on cardiac MRI and 15.6% on abdominal CT versus UNETR, reaching the top compared mean on AMOS.

desk verdict A sensible incremental architecture paper whose headline gains over UNETR are not apples-to-apples, and whose AMOS 'SOTA' gap is within noise; worth refereeing but needs protocol fixes. read the letter →

arxiv 2501.07017 v2 pith:MVJM4L32 submitted 2025-01-13 cs.CV cs.AI

Xuhui Guo , Tanmoy Dam , Rohan Dhamdhere , Gourav Modanwal , Anant Madabhushi This is my paper · ORCID

classification cs.CVcs.AI

keywords 3DmedicalimagesegmentationVision-LSTMChebyshevKANUNETRU-Netencoder-decoderACDCcardiacMRIAMOS2022abdominalCTlong-rangedependencymodeling

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 a 3D U-Net segmentation model with a Vision-LSTM encoder and Chebyshev-KAN projection layers outperforms the ViT-based UNETR it modifies, reporting mean Dice rising from 85.34% to 91.59% on the ACDC cardiac MRI benchmark and from 76.59% to 88.57% on the AMOS 2022 post-challenge Task 2 abdominal CT benchmark. On AMOS, the model reaches the highest mean Dice among compared methods. The reason to care is that Vision-LSTM models long-range dependencies with linear rather than quadratic complexity, while Chebyshev KAN sharpens boundary detail and curbs over-segmentation. The paper also reports an ablation showing Chebyshev beats MLP, B-spline, and Gaussian RBF as the KAN univariate function in a lightweight segmentation model.

What carries the argument

The load-bearing objects are the ViL block pair and the Chebyshev KAN layer. A ViL block alternates forward and backward mLSTM layers, whose matrix-valued memory cell lets the encoder accumulate long-range spatial context over patch tokens with linear cost in sequence length. The Chebyshev KAN layer replaces the MLP in the ViL up and down projections: Chebyshev polynomials $T_0(x)=1$, $T_1(x)=x$, and $T_m(x)=2xT_{m-1}(x)-T_{m-2}(x)$ are combined with a learnable coefficient tensor $C$ of shape $K \times O \times (d+1)$ by Einstein summation to produce the projected feature map. Together they give the encoder long-range memory and the projection paths higher nonlinear capacity.

What would settle it

Train the UNETR baseline and UNETVL with identical data splits, preprocessing, augmentation, batch size, optimizer, and epoch counts, then compare per-fold mean Dice with standard deviations. If the reported 7.3% (ACDC) and 15.6% (AMOS) gaps shrink to roughly 2% or less, the claimed advantage is a protocol artifact rather than an architectural effect.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is that a U-Net built from Vision-LSTM encoder blocks, with Chebyshev-KAN projections in place of MLP projections, segments 3D medical volumes more accurately than the ViT-based UNETR it modifies: mean Dice rises from 85.34% to 91.59% on ACDC and from 76.59% to 88.57% on AMOS 2022 post-challenge Task 2, the highest mean among compared methods on the latter. The authors attribute this to bidirectional mLSTM memory capturing long-range spatial context with lower computational cost than self-attention, and to Chebyshev polynomials giving the projection layers stronger nonlinear approximation. They position the model as an efficiency-accuracy alternative to both convolutional and transformer encoders, with the network architecture as the only variable in an otherwise standardized pipeline.

Load-bearing premise

The load-bearing assumption is that the published UNETR and other comparison scores were obtained under the same preprocessing, training, and evaluation protocol as UNETVL, so the reported Dice gains reflect the architecture rather than the pipeline.

Editorial extensions

If this is right

A ViL-based encoder gives a transformer-free way to model long-range spatial dependencies in 3D volumes with linear rather than quadratic complexity.
Chebyshev KAN projections can be swapped into ViL up and down projection layers with a one-line change, raising Dice in the reported settings.
On the AMOS 2022 post-challenge Task 2 dataset, the architecture reaches the highest mean Dice among compared methods, supporting abdominal multi-organ segmentation.
The ablations show that raising the latent dimension $K$ improves accuracy at higher parameter counts, giving an explicit accuracy-cost trade-off for deployment.

Reading between the lines

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

Because several comparison-table baselines were taken from a prior study rather than rerun here, the 7.3% and 15.6% gains should be read as provisional until a same-protocol head-to-head confirms them.
The Chebyshev-versus-MLP advantage was isolated in a lightweight segmentation model on ACDC; whether it transfers to other organs, modalities, or full UNETVL scales is untested.
ViL's linear complexity suggests the encoder could be pushed to larger patch counts or higher input resolutions than the tested settings.
The qualitative over-segmentation reduction could be made quantitative with surface Dice or Hausdorff distance, which would strengthen the clinical claim.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

The paper proposes UNETVL, a UNETR variant that swaps the ViT encoder for Vision-LSTM blocks and replaces the MLP projections with a Chebyshev KAN. That's a straightforward, reasonable combination, and the paper is honestly written around it. The work is incremental, but not every paper needs to be a breakthrough to be useful.

What it does well: the implementation inside nnU-Net is the right way to control for preprocessing and augmentation, and the comparisons to SwinUNETR, nnFormer, and others that were also run in the nnU-Net-revisited framework are the credible part of the evaluation. The univariate-function ablation on U-KAN is a sensible sanity check, even if it's on a different architecture. Code is promised.

The soft spots are real and they matter. First, the headline deltas — 7.3% on ACDC, 15.6% on AMOS — compare against UNETR numbers taken from the original UNETR paper, not from a re-run inside nnU-Net. The claim in Section 3 that the setup 'ensures a fair comparison with the network architecture as the sole variable' is simply not true for that baseline. Second, the AMOS 'highest mean Dice' claim is 88.57 vs CoTr's 88.02, a 0.55-point gap with no error bars or significance test. That is within the usual spread of 5-fold CV on these benchmarks. Third, the main ablation table (Table 2) is run on a single fold after 100 epochs, which doesn't match the reported final results in Table 1, so it doesn't cleanly demonstrate that KAN is the cause of the gain. Fourth, the choice of Chebyshev was made on ACDC via the U-KAN experiments, and then evaluated on ACDC — a mild selection bias, not a fatal one.

None of this kills the architecture. The within-pipeline comparisons suggest ViL is a competitive encoder, and the KAN swap is a plausible tweak. But the public claims overstate what is established. The paper deserves a serious referee, provided the referee pushes for a re-run of UNETR under identical conditions, error bars on all headline numbers, and a clearer ablation protocol. If the authors fix those, this becomes a solid contribution to the incremental-architecture literature.

My take: send it to review, but expect major revision.

Referee Report

5 major / 6 minor

Summary. The paper proposes UNETVL, a 3D medical image segmentation architecture that replaces the ViT encoder of UNETR with Vision-LSTM (ViL) blocks and injects Chebyshev KAN layers into the up- and down-projection paths. The method is implemented inside the nnU-Net framework and evaluated on the ACDC and AMOS 2022 post-challenge Task 2 datasets. The authors report mean Dice scores of 91.59 on ACDC and 88.57 on AMOS, claim improvements of 7.3% and 15.6% over UNETR, and state that the AMOS result is the highest among the compared methods. Ablation studies on the ACDC dataset are used to justify the Chebyshev KAN choice and the latent dimension.

Significance. If the reported results hold under uniform evaluation, the paper would present a strong and practical 3D segmentation backbone, with the code release and the nnU-Net-based implementation being concrete strengths. However, the central quantitative claims are currently not established because the UNETR baseline is taken from an external study with a different protocol, and the AMOS state-of-the-art claim rests on a 0.55-point margin without any uncertainty quantification. The architecture itself is plausible, and the within-pipeline comparisons against SwinUNETR and other modern baselines are informative, but the headline deltas and the SOTA designation need more rigorous support before the paper can be accepted.

major comments (5)

[Section 3, Implementation Details; Table 1] The claim that the nnU-Net framework 'ensures a fair comparison ... with the network architecture as the sole variable' is not true for the UNETR baseline. Table 1's caption states that all results except UNETR and SAM3D come from the nnU-Net revisited study, while UNETVL is trained using nnU-Net preprocessing, augmentation, oversampling, and 5-fold evaluation. The UNETR Dice scores (85.34 on ACDC, 76.59 on AMOS) are legacy numbers from the original UNETR paper, which used a different protocol. Consequently, the 7.3% and 15.6% improvements advertised in the abstract are cross-protocol comparisons and may be artifacts of preprocessing or training differences rather than architectural superiority. The authors should rerun UNETR inside the nnU-Net pipeline, or explicitly state that the UNETR comparison is not apples-to-apples and remove the quantitative improvement claims from the abstract.
[Section 4, Table 1] The claim that UNETVL 'achieves the highest mean Dice score on the AMOS 2022 post-challenge Task 2 dataset' is not statistically supported. The margin over CoTr is only 0.55 percentage points (88.57 vs. 88.02), yet no per-fold results, standard deviations, confidence intervals, or significance tests are reported anywhere in the paper. With 5-fold cross-validation, a difference of this size can easily arise from fold-to-fold variability. The authors should report per-fold scores and some measure of uncertainty before making a SOTA claim.
[Section 4, Table 2] The ablation table is presented without clarifying that it uses a different evaluation protocol than the final results: the caption says 'test on ACDC cross validation fold 1 after 100 epochs', whereas Table 1 reports 5-fold cross-validation results. As a result, the reader cannot directly compare the ablation numbers (e.g., 89.79 for w/ KAN, K=384 in Table 2) with the main result (91.59 in Table 1), and the conclusion that 'increasing the latent dimension results in better segmentation' is only supported on a single fold after 100 epochs. The authors should either run the full ablation over all 5 folds and the full training schedule, or explicitly state that the ablation is preliminary and only indicative.
[Section 4, Choice of Univariate Functions for KAN] The selection of Chebyshev polynomials as the univariate function in the KAN layer is made using 5-fold cross-validation on the ACDC dataset, and the final UNETVL model is then evaluated on the same ACDC benchmark. This creates a selection bias: the function choice is tuned to ACDC and the reported ACDC performance of the full model is not an unbiased estimate. Moreover, the difference between Chebyshev (86.12) and MLP (85.88) in Table 3 is only 0.24 percentage points, and no error bars are provided. The paper's claim that Chebyshev 'consistently demonstrated superior performance' is therefore not firmly established. The authors should report the variance of these estimates, evaluate the function choice on a held-out dataset (e.g., AMOS), or temper the claim.
[Throughout; Sections 3 and 4] No error bars, standard deviations, or statistical significance tests are provided for any of the main segmentation results. Given that the paper's central claims include a 7.3% improvement on ACDC and a 0.55-point SOTA margin on AMOS, the absence of uncertainty quantification is a load-bearing omission. The authors should report per-fold Dice scores and at least the standard deviation across folds for all methods they compare, especially for their own model.

minor comments (6)

[Section 2.3, Eq. (1)] The tensor T is defined as having shape R^{N x K x (degree+1)}, and the summation is written over i and j, but the index j is not explicitly defined in the text as the polynomial-degree index; please add a sentence for clarity.
[Section 2.2] The description of the mLSTM block would benefit from equations or a precise pointer to the xLSTM paper; as written, the memory-cell mechanism is described only verbally, which makes the architecture hard to reproduce from the paper alone.
[Figure 1] The right panel of Figure 1 is low-contrast and the labels 'Even', 'False', 'Flip', and the KAN projection annotations are difficult to read; please enlarge the figure or split it into separate panels.
[Table 3] There is a formatting issue: the caption appears as 'Table 3 .' and the table body is not well spaced; also, the rows for MLP, B-Spline, Gaussian RBF, and Chebyshev should be visually separated.
[Section 5, Discussion] The statement that UNETVL 'significantly mitigates over-segmentation' is not quantified; Figure 2 shows qualitative examples, but a quantitative measure such as boundary F1, surface Dice, or a per-structure over-segmentation metric would support the claim.
[Throughout] The paper uses both 'Kolmogorov–Arnold Networks (KAN)' and 'Kolmogorov-Arnold Network' inconsistently; please unify the terminology and check singular/plural usage.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the architecture's claims are empirical and its components are cited from external work, not derived from the target result.

full rationale

UNETVL is an empirical architecture paper. The ViL encoder and Chebyshev KAN layers are adopted from external prior work (references [3,4,5,9]) and are not derived from the ACDC or AMOS Dice scores. The reported Dice values are measured outcomes, not quantities solved for from the definition of the model. Table 1's UNETR baseline is imported from the original UNETR study rather than re-run inside the nnU-Net pipeline, and the Chebyshev activation is selected via ACDC cross-validation before final ACDC reporting; both are protocol and selection concerns that could affect the validity of the comparison, but neither makes a prediction equivalent to its input by construction. There are no load-bearing self-citations, no fitted parameter renamed as a prediction, and no uniqueness theorem imported from the authors. Under the hard rules, such concerns belong to correctness risk, not circularity, so the circularity score is 0.

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

The central claim is empirical and rests on standard deep learning components. The main free parameters are architecture choices (embedding dimension, polynomial degree, ViL depth) that are not derived from theory. The comparison relies on assumptions about baseline comparability and the transferability of the Chebyshev choice from U-KAN to UNETVL. No fundamentally new theoretical entities are introduced.

free parameters (3)

Embedding dimension K = 384
The final model uses K=384. Ablations in Table 2 test K=192 and K=384, and the larger dimension gives higher Dice but more parameters. This is a hand-chosen architecture hyperparameter.
Chebyshev polynomial degree = Not reported
The Chebyshev KAN layer has a polynomial degree that determines the number of coefficients; the paper does not state the degree used, so the exact parameter is unspecified.
Number of ViL block pairs per stage = Not reported
The architecture diagram shows stages but the text does not specify how many ViL block pairs are used; this affects model capacity and is set by the authors.

assumptions (3)

domain assumption Vision-LSTM as described by Alkin et al. (2024) provides a valid and effective backbone when adapted to 3D patch tokens.
The paper adopts ViL without modification except for adding KAN projections; it assumes the mLSTM blocks process 3D tokens as intended.
domain assumption The nnU-Net framework's preprocessing, augmentation, and training schedule are appropriate for all compared architectures.
UNETVL is implemented in nnU-Net, and baseline results are drawn from the nnU-Net revisited paper; the fairness of the comparison depends on this assumption.
ad hoc to paper Chebyshev polynomial basis functions can represent the needed feature transformations at least as well as MLPs.
The paper provides no theoretical justification; it relies on the empirical ablation in Table 3 (on a different U-KAN architecture) to claim Chebyshev is superior.

how reviews work

0 comments

Cite this review

Pith. "Pith review of UNetVL: Enhancing 3D Medical Image Segmentation with Chebyshev KAN Powered Vision-LSTM." pith.science (2026). https://pith.science/paper/MVJM4L32

@misc{pith2026250107017,
  author       = {Pith},
  title        = {Pith review of: UNetVL: Enhancing 3D Medical Image Segmentation with Chebyshev KAN Powered Vision-LSTM},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MVJM4L32}},
  note         = {Machine review of arXiv:2501.07017}
}

read the original abstract

3D medical image segmentation has progressed considerably due to Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs), yet these methods struggle to balance long-range dependency acquisition with computational efficiency. To address this challenge, we propose UNETVL (U-Net Vision-LSTM), a novel architecture that leverages recent advancements in temporal information processing. UNETVL incorporates Vision-LSTM (ViL) for improved scalability and memory functions, alongside an efficient Chebyshev Kolmogorov-Arnold Networks (KAN) to handle complex and long-range dependency patterns more effectively. We validated our method on the ACDC and AMOS2022 (post challenge Task 2) benchmark datasets, showing a significant improvement in mean Dice score compared to recent state-of-the-art approaches, especially over its predecessor, UNETR, with increases of 7.3% on ACDC and 15.6% on AMOS, respectively. Extensive ablation studies were conducted to demonstrate the impact of each component in UNETVL, providing a comprehensive understanding of its architecture. Our code is available at https://github.com/tgrex6/UNETVL, facilitating further research and applications in this domain.

Figures

Figures reproduced from arXiv: 2501.07017 by the authors.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 14 canonical work pages

[1]

INTRODUCTION The incorporation of artificial intelligence (AI) into medi- cal image process has profoundly revolutionized patient care, providing a more efficient and accessible approach to im- proving diagnostic precision and prognostic outcomes. Accu- rate segmentation, in particular, acts as a fundamental step in biomedical image analysis, delineating ...

work page Pith review arXiv 2025
[2]

Architecture Overview We present the overall architecture of our UNETVL model in Fig

METHODOLOGY 2.1. Architecture Overview We present the overall architecture of our UNETVL model in Fig. 1. Building upon the UNETR structure [1], UNETVL employs an encoder-decoder network that efficiently captures both local features and long-range contexts. The model pro- cesses a 3D oversampled instance with resolution (H, W, D) and C input channels, app...

work page
[3]

EXPERIMENTS Datasets. We evaluate the effectiveness of our approach on two public 3D segmentation benchmarks, both involving mul- tiple structures or multiple organs: the Automated Cardiac Linear Projection 1 2 3 4 5 6 7 8 9 Embedded Patches + ViL Block Pairs Stage 2 ViL Block Pairs Stage 1 ViL Block Pairs Stage 3 ViL Block Pairs Stage 4 UNETR Decoder Lay...

work page 2022
[4]

As shown in Table 1, when comparing the performance of the ViL architecture with Chebyshev KAN to the UNETR baseline, a signifi- cant improvement is observed

RESULTS Evaluation on ACDC and AMOS. As shown in Table 1, when comparing the performance of the ViL architecture with Chebyshev KAN to the UNETR baseline, a signifi- cant improvement is observed. On the ACDC dataset, the Dice score improves by approximately 7.3% (from 85.34% in UNETR to 91.59% in UNETVL with KAN). Similarly, for the AMOS dataset, the Dice...

work page 2022
[5]

DISCUSSION AND CONCLUSION The proposed UNETVL architecture, enhanced by Cheby- shev KAN, demonstrates competitive performance in 3D medical image segmentation tasks. While the Dice score may not be the absolute highest, the model significantly mitigates over-segmentation, capturing finer anatomical de- Model Configuration #Params (M) Mean Dice Score (%) w...

work page
[6]

[6] and Ji et al

COMPLIANCE WITH ETHICAL STANDARDS This research study was conducted retrospectively using hu- man subject data made available in open access by Bernard et al. [6] and Ji et al. [7]. Ethical approval was not required as confirmed by the license attached with the open access data

work page
[7]

Unetr: Transformers for 3d medical image segmentation,

Ali Hatamizadeh, Yucheng Tang, Vishwesh Nath, Dong Yang, Andriy Myronenko, Bennett Landman, Holger R Roth, and Daguang Xu, “Unetr: Transformers for 3d medical image segmentation,” in Proceedings of the IEEE/CVF winter conference on applications of com- puter vision, 2022, pp. 574–584

2022
[8]

U-net: Convolutional networks for biomedical im- age segmentation,

Olaf Ronneberger, Philipp Fischer, and Thomas Brox, “U-net: Convolutional networks for biomedical im- age segmentation,” in Medical image computing and computer-assisted intervention–MICCAI 2015: 18th in- ternational conference, Munich, Germany, October 5-9, 2015, proceedings, part III 18. Springer, 2015, pp. 234– 241

2015

Show all 22 references

[9]

xlstm: Extended long short-term memory,

Maximilian Beck, Korbinian P ¨oppel, Markus Spanring, Andreas Auer, Oleksandra Prudnikova, Michael Kopp, G¨unter Klambauer, Johannes Brandstetter, and Sepp Hochreiter, “xlstm: Extended long short-term memory,” arXiv preprint arXiv:2405.04517, 2024

2024 arXiv
[10]

Vision- lstm: xlstm as generic vision backbone,

Benedikt Alkin, Maximilian Beck, Korbinian P ¨oppel, Sepp Hochreiter, and Johannes Brandstetter, “Vision- lstm: xlstm as generic vision backbone,” arXiv preprint arXiv:2406.04303, 2024

2024 arXiv
[11]

Kan: Kolmogorov-arnold net- works,

Ziming Liu, Yixuan Wang, Sachin Vaidya, Fabian Ruehle, James Halverson, Marin Solja ˇci´c, Thomas Y Hou, and Max Tegmark, “Kan: Kolmogorov-arnold net- works,” arXiv preprint arXiv:2404.19756, 2024

2024 arXiv
[12]

Deep learning techniques for automatic mri cardiac multi-structures segmentation and diagnosis: is the problem solved?,

Olivier Bernard, Alain Lalande, Clement Zotti, Fred- erick Cervenansky, Xin Yang, Pheng-Ann Heng, Irem Cetin, Karim Lekadir, Oscar Camara, Miguel An- gel Gonzalez Ballester, et al., “Deep learning techniques for automatic mri cardiac multi-structures segmentation and diagnosis...

2018
[13]

Amos: A large-scale abdominal multi-organ benchmark for versa- tile medical image segmentation,

Yuanfeng Ji, Haotian Bai, Chongjian GE, Jie Yang, Ye Zhu, Ruimao Zhang, Zhen Li, Lingyan Zhanng, Wanling Ma, Xiang Wan, and Ping Luo, “Amos: A large-scale abdominal multi-organ benchmark for versa- tile medical image segmentation,” in Advances in Neu- ral Information Processin...

2022
[14]

nnu-net revisited: A call for rig- orous validation in 3d medical image segmentation,

Fabian Isensee, Tassilo Wald, Constantin Ulrich, Michael Baumgartner, Saikat Roy, Klaus Maier-Hein, and Paul F Jaeger, “nnu-net revisited: A call for rig- orous validation in 3d medical image segmentation,” in International Conference on Medical Image Com- puting and Computer-...

2024
[15]

Chebyshev polynomial-based kolmogorov-arnold networks: An efficient archi- tecture for nonlinear function approximation,

Sidharth SS, “Chebyshev polynomial-based kolmogorov-arnold networks: An efficient archi- tecture for nonlinear function approximation,” arXiv preprint arXiv:2405.07200, 2024

2024 arXiv
[16]

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

Fabian Isensee, Paul F Jaeger, Simon AA Kohl, Jens Petersen, and Klaus 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

2021
[17]

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

Ali Hatamizadeh, Vishwesh Nath, Yucheng Tang, Dong Yang, Holger R Roth, and Daguang Xu, “Swin unetr: Swin transformers for semantic segmentation of brain tumors in mri images,” in International MICCAI brain- lesion workshop. Springer, 2021, pp. 272–284

2021
[18]

Swinunetr- v2: Stronger swin transformers with stagewise convo- lutions for 3d medical image segmentation,

Yufan He, Vishwesh Nath, Dong Yang, Yucheng Tang, Andriy Myronenko, and Daguang Xu, “Swinunetr- v2: Stronger swin transformers with stagewise convo- lutions for 3d medical image segmentation,” in In- ternational Conference on Medical Image Computing and Computer-Assisted Inter...

2023
[19]

nnformer: V olumetric medical image segmenta- tion via a 3d transformer,

Hong-Yu Zhou, Jiansen Guo, Yinghao Zhang, Xi- aoguang Han, Lequan Yu, Liansheng Wang, and Yizhou Yu, “nnformer: V olumetric medical image segmenta- tion via a 3d transformer,” IEEE Transactions on Image Processing, vol. 32, pp. 4036–4045, 2023

2023
[20]

Cotr: Efficiently bridging cnn and transformer for 3d medical image segmentation,

Yutong Xie, Jianpeng Zhang, Chunhua Shen, and Yong Xia, “Cotr: Efficiently bridging cnn and transformer for 3d medical image segmentation,” in Medical Im- age Computing and Computer Assisted Intervention– MICCAI 2021: 24th International Conference, Stras- bourg, France, Septem...

2021
[21]

Sam3d: Segment any- thing model in volumetric medical images,

Nhat-Tan Bui, Dinh-Hieu Hoang, Minh-Triet Tran, Gi- anfranco Doretto, Donald Adjeroh, Brijesh Patel, Ara- binda Choudhary, and Ngan Le, “Sam3d: Segment any- thing model in volumetric medical images,” in 2024 IEEE International Symposium on Biomedical Imaging (ISBI), 2024, pp. 1–4

2024
[22]

U-kan makes strong backbone for medical image seg- mentation and generation,

Chenxin Li, Xinyu Liu, Wuyang Li, Cheng Wang, Hengyu Liu, Yifan Liu, Zhen Chen, and Yixuan Yuan, “U-kan makes strong backbone for medical image seg- mentation and generation,” 2024

2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

INTRODUCTION The incorporation of artificial intelligence (AI) into medi- cal image process has profoundly revolutionized patient care, providing a more efficient and accessible approach to im- proving diagnostic precision and prognostic outcomes. Accu- rate segmentation, in particular, acts as a fundamental step in biomedical image analysis, delineating ...

work page Pith review arXiv 2025

[2] [2]

Architecture Overview We present the overall architecture of our UNETVL model in Fig

METHODOLOGY 2.1. Architecture Overview We present the overall architecture of our UNETVL model in Fig. 1. Building upon the UNETR structure [1], UNETVL employs an encoder-decoder network that efficiently captures both local features and long-range contexts. The model pro- cesses a 3D oversampled instance with resolution (H, W, D) and C input channels, app...

work page

[3] [3]

EXPERIMENTS Datasets. We evaluate the effectiveness of our approach on two public 3D segmentation benchmarks, both involving mul- tiple structures or multiple organs: the Automated Cardiac Linear Projection 1 2 3 4 5 6 7 8 9 Embedded Patches + ViL Block Pairs Stage 2 ViL Block Pairs Stage 1 ViL Block Pairs Stage 3 ViL Block Pairs Stage 4 UNETR Decoder Lay...

work page 2022

[4] [4]

As shown in Table 1, when comparing the performance of the ViL architecture with Chebyshev KAN to the UNETR baseline, a signifi- cant improvement is observed

RESULTS Evaluation on ACDC and AMOS. As shown in Table 1, when comparing the performance of the ViL architecture with Chebyshev KAN to the UNETR baseline, a signifi- cant improvement is observed. On the ACDC dataset, the Dice score improves by approximately 7.3% (from 85.34% in UNETR to 91.59% in UNETVL with KAN). Similarly, for the AMOS dataset, the Dice...

work page 2022

[5] [5]

DISCUSSION AND CONCLUSION The proposed UNETVL architecture, enhanced by Cheby- shev KAN, demonstrates competitive performance in 3D medical image segmentation tasks. While the Dice score may not be the absolute highest, the model significantly mitigates over-segmentation, capturing finer anatomical de- Model Configuration #Params (M) Mean Dice Score (%) w...

work page

[6] [6]

[6] and Ji et al

COMPLIANCE WITH ETHICAL STANDARDS This research study was conducted retrospectively using hu- man subject data made available in open access by Bernard et al. [6] and Ji et al. [7]. Ethical approval was not required as confirmed by the license attached with the open access data

work page

[7] [7]

Unetr: Transformers for 3d medical image segmentation,

Ali Hatamizadeh, Yucheng Tang, Vishwesh Nath, Dong Yang, Andriy Myronenko, Bennett Landman, Holger R Roth, and Daguang Xu, “Unetr: Transformers for 3d medical image segmentation,” in Proceedings of the IEEE/CVF winter conference on applications of com- puter vision, 2022, pp. 574–584

2022

[8] [8]

U-net: Convolutional networks for biomedical im- age segmentation,

Olaf Ronneberger, Philipp Fischer, and Thomas Brox, “U-net: Convolutional networks for biomedical im- age segmentation,” in Medical image computing and computer-assisted intervention–MICCAI 2015: 18th in- ternational conference, Munich, Germany, October 5-9, 2015, proceedings, part III 18. Springer, 2015, pp. 234– 241

2015