Pith. sign in

REVIEW 5 major objections 6 minor 46 references

T-MPEDNet: Unveiling the Synergy of Transformer-aware Multiscale Progressive Encoder-Decoder Network with Feature Recalibration for Tumor and Liver Segmentation

T0 review · 5 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read T-MPEDNet claims state-of-the-art CT liver and tumor segmentation, with 97.6% and 89.1% Dice on LiTS and 98.3% and 83.3% on 3DIRCADb.

desk verdict Coherent architecture and an honest ablation, but the 'beats all twelve' claim rests on an uncontrolled comparison: split may be slice-level and baselines are borrowed. read the letter →

arxiv 2507.19590 v1 pith:YJMCDU55 submitted 2025-07-25 cs.CV

Chandravardhan Singh Raghaw , Jasmer Singh Sanjotra , Mohammad Zia Ur Rehman , Shubhi Bansal , Shahid Shafi Dar , Nagendra Kumar This is my paper · ORCID
classification cs.CV
keywords livertumorsegmentationcomputedtomographyprogressiveencoder-decodertransformerattentionmulti-scaleatrousconvolutionchannelrecalibrationmorphologicalboundaryrefinementDicesimilaritycoefficient
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 aims to show that a new 2D neural network, T-MPEDNet, can segment the liver and liver tumors in CT scans more accurately than twelve published methods. It reports Dice similarity coefficients of 97.6% for liver and 89.1% for tumor on the LiTS benchmark, and 98.3% and 83.3% on 3DIRCADb, with the largest reported gains on tumors, the harder class. The architecture combines a progressive five-stage encoder-decoder with channel-wise feature recalibration, a transformer-inspired attention module, multi-scale dilated convolutions, and a morphological boundary-refinement step. If the results hold, the network would be an effective automated contouring tool for liver diagnosis and treatment planning.

What carries the argument

The load-bearing idea is a progressive encoder-decoder built from five Adaptive Feature Extraction (AdaFEx) stages, joined by skip connections that go through a Compressive Channel Recalibration (CCR) module before each decoder stage. A Dynamic Contextual Attention (DCA) module applies dynamic convolutions before multi-head self-attention to capture long-range spatial context, a Multi-Scale Atrous Spatial (MSAS) module uses parallel dilated convolutions with dilation rates 1, 4, 8, and 12 to span local and global scales, and a Morphological Boundary Refinement (MBR) post-processor erodes the output mask to isolate and sharpen boundary pixels.

What would settle it

Run T-MPEDNet and all twelve compared methods on the same 80/10/10 split of LiTS and 3DIRCADb, with identical preprocessing (HU windowing, CLAHE, z-score normalization, and 256 by 256 resizing) and the same optimizer and augmentation, then compare Dice on the identical test slices; if the margins shrink below the reported 0.4 to 1.2 points or reverse sign, the central claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that T-MPEDNet surpasses all twelve compared methods on both public benchmarks. The network couples a five-stage progressive encoder-decoder whose skip connections pass through a compressive channel recalibration unit with a transformer-inspired dynamic contextual attention module and a multi-scale atrous spatial module, then erodes the predicted mask to sharpen boundaries. On LiTS it reports Dice similarity coefficients of 97.6% and 89.1% for liver and tumor; on 3DIRCADb it reports 98.3% and 83.3%. Ablations on LiTS attribute the largest single contributions to channel recalibration and dynamic attention.

Load-bearing premise

The result depends on the baseline methods being evaluated under the same data split and preprocessing as the new network; if they were not, the reported advantages could come from the test setup rather than from the architecture.

Editorial extensions

If this is right

  • On LiTS, the full model reports 97.6% Dice for liver and 89.1% for tumor; on 3DIRCADb it reports 98.3% and 83.3%.
  • The ablation study shows that the compressive channel recalibration module is the largest single contributor, with its removal dropping liver Dice by 4.1 points and tumor Dice by 7.7 points on LiTS.
  • The dynamic contextual attention module contributes the second-largest margin, with its removal dropping liver Dice by 2.2 points and tumor Dice by 5.2 points.
  • The morphological boundary refinement adds a smaller margin (0.2 liver, 0.5 tumor Dice on LiTS) and works as a post-processing step, so it could be reused without retraining the network.

Reading between the lines

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

  • A controlled re-run with all twelve baselines retrained under identical splits and preprocessing would test whether the reported 0.4 to 1.2 point margins are due to the architecture; the paper does not report such a re-run.
  • Because boundary refinement is a generic morphological operation, it may transfer to other low-contrast organ segmentation tasks, such as pancreas or kidney tumors, without architectural changes.
  • The largest gains come from modules that reweight channels and capture long-range context, so a similar combination could help small-lesion segmentation in other CT modalities; testing this would require new experiments.
  • Training on 2D slices at 256 by 256 resolution leaves inter-slice context unused, so a 3D or higher-resolution variant is a natural next step that could push tumor Dice further; the paper lists this direction as future work.
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

5 major / 6 minor

Summary. The paper proposes T-MPEDNet, a 2D progressive encoder-decoder for liver and tumor segmentation in CT volumes. The architecture combines Adaptive Feature Extraction (AdaFEx) stages, a Compressive Channel Recalibration (CCR) module, a Transformer-inspired Dynamic Contextual Attention (DCA) module, a Multi-Scale Atrous Spatial (MSAS) module, and a Morphological Boundary Refinement (MBR) postprocessing step. The authors report Dice scores of 97.6/89.1 for liver/tumor on a self-selected LiTS split and 98.3/83.3 on 3DIRCADb, and on this basis claim superiority over twelve prior methods. A five-row ablation on LiTS attributes performance gains to the individual components.

Significance. If the empirical claims are supported, the paper would provide a competitive 2D architecture for an important clinical segmentation task, with a clear module breakdown and ablations. The architectural components are described in enough detail to be reimplemented, and the use of two public datasets is appropriate. The main value would lie in the reported benchmark results. However, the significance is currently limited by the uncontrolled evaluation protocol: the data split appears to be at slice level rather than patient level, the comparison baselines are taken from disparate published protocols, and no uncertainty quantification is provided. These issues directly affect the central claim of outperforming all twelve compared methods.

major comments (5)
  1. [Section 4.1.1, Table 2] The data split is described in Section 4.1.1 as randomly splitting CT images into 80% training, 10% validation, and 10% testing, and Table 2 reports slice counts (54,065/5,001/4,573 for LiTS and 2,258/282/282 for 3DIRCADb). This indicates a slice-level split rather than a patient-level or volume-level split. Adjacent axial slices from the same CT volume are highly correlated, so slices of the same liver and tumor can appear in both training and test sets, inflating the reported Dice scores. A patient-level split with case-level holdout, together with per-case metrics, is needed before the reported superiority can be accepted.
  2. [Table 3, Section 4.2.1] The comparison with twelve state-of-the-art methods relies on Dice scores taken from the original papers, with no evidence that those methods were run under the same preprocessing (HU windowing, CLAHE, z-score normalization), the same 256x256 resampling, the same slice selection, or the same train/validation/test partition. The reported margins over the second-best methods are only 0.4 to 1.2 DSC points, which is within the range of split-to-split and protocol-to-protocol variation. Without a controlled comparison under a common protocol, the claim that T-MPEDNet surpasses all twelve methods is not supported.
  3. [Section 4.2.1, Tables 3 and 4] No error bars, standard deviations, number of independent runs, or significance tests are reported. The central comparison is based on single Dice values, and the smallest claimed margins are around 0.4 DSC points. At minimum, the authors should report mean and standard deviation over several runs with different seeds, and ideally paired statistical tests for the differences against the strongest baselines.
  4. [Section 4.2.2, Table 4] The ablation interpretation in the text does not match the table. The sentence 'Removing the DCA module ... declines the liver DSC by 2.2% and tumor DSC by 5.2%' compares the full model with the row that removes DCA, MSAS, and MBR simultaneously, not DCA alone. Similarly, the reported 1.7% and 2.7% drops for removing MSAS are not the differences between consecutive rows. One-at-a-time ablations, with a single module removed while all others remain, are necessary to attribute performance changes to individual components.
  5. [Section 3.2.3, Eq. (12), Fig. 8] The Morphological Boundary Refinement module is described as producing a boundary mask BMask by subtracting an eroded mask from the input mask, but the manuscript never specifies how BMask is combined with the network's output to obtain the final segmentation. Without an explicit final-mask equation, the reported 0.2% and 0.5% Dice improvements from MBR in Table 4 cannot be reproduced or verified. Please state the exact postprocessing rule and clarify whether the reported scores include this step.
minor comments (6)
  1. [Eq. (4)] The summation in Eq. (4) runs to H×C, while the text states the spatial dimensions are H×W; the summation index j is also undefined. Please correct the upper limit and define the index.
  2. [Eq. (6)] Eq. (6) writes f_j^CCR = a_k ⊙ z_k, mixing indices j and k. This should presumably be f_k^CCR = a_k ⊙ z_k for each channel k.
  3. [Eq. (5)] The notation σ^Sigmoid and α^ReLU is unconventional and hard to read; standard mathematical notation for the sigmoid and ReLU functions should be used.
  4. [Eq. (11)] Eq. (11) uses plus signs between the four atrous-convolution terms, although the text says the multi-scale features are concatenated. Please use a concatenation symbol or clarify the operation.
  5. [Table 2] There are typographical issues in the column headers, including 'V oxel spacing' with a space, and the Z-axis spacing range is written as '01.6 - 4.05 mm' instead of 1.6-4.05 mm.
  6. [Title page and Abstract] The abstract appears twice at the beginning of the document, and the arXiv header says 'Preprint submitted to Expert Systems with Applications' while the highlight page says the paper is accepted in Biomedical Signal Processing and Control. Please align these statements and remove the duplication.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's claims are empirical benchmark comparisons against external public datasets, not derivations that reduce to fitted values or self-citations.

full rationale

I walked the claimed derivation chain from the T-MPEDNet architecture (Sections 3.2.1-3.2.3) through the experimental evaluation (Section 4) to the stated superiority claims. The central claims are (a) that the proposed network architecture achieves high DSC values on LiTS and 3DIRCADb, and (b) that it outperforms twelve prior methods. These are empirical statements evaluated against externally labeled ground-truth masks from public benchmarks; no equation in the paper defines a predicted quantity in terms of the very data it is supposed to predict. The components (AdaFEx, CCR, DCA, MSAS, MBR) are described by explicit forward computations (Eqs. 2-12), and their contributions are assessed through ablation (Table 4), which is internal consistency checking rather than circular reasoning. The hyperparameter gamma=4 is said to be empirically chosen (Section 3.2.2.2), which is standard tuning and does not make the benchmark results circular. The only reference that overlaps with the present authors is [40], cited in a general survey sentence about multi-scale techniques in other domains; it is not load-bearing for any architectural premise or uniqueness claim. There is no self-citation chain invoked to forbid alternative designs, no imported uniqueness theorem, and no ansatz smuggled in via citation. The concerns raised by a skeptical reader are about experimental validity: the train/validation/test split appears to be slice-level rather than patient-level (Table 2), and baseline numbers in Table 3 may come from heterogeneous protocols. Those are legitimate external-validity or reproducibility concerns, but they are not circularity: an uncontrolled comparison can be weak evidence without the result being equivalent to its own inputs by construction. Under the scoring rubric, the correct finding is no significant circularity, score 0.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard empirical assumptions: 2D slices represent 3D volumes, the random split is a fair benchmark, baseline numbers are comparable, and the morphological refinement improves boundaries. The paper also selects several hyperparameters by hand or by validation tuning, notably gamma=4 and dilation rates.

free parameters (7)
  • Initial channel count C_f = 16
    Sets model width in every AdaFEx stage; chosen by hand with no sensitivity analysis.
  • CCR compression ratio gamma = 4
    Paper states 'We empirically found that gamma=4 yields optimal performance' (Section 3.2.2(ii)); tuned on validation data.
  • MSAS dilation rates = text says 1,4,8,12; Table 1 says 4,8,12
    Design choice for receptive fields; the inconsistency between text and table is unresolved.
  • Number of attention heads H = not specified
    MHDCA partitions features into H subspaces, but H is never reported in Section 3.2.2(iii).
  • Preprocessing parameters = resize 256x256, HU window (-250,200), CLAHE, z-score
    These choices determine which image information the network sees and are not varied.
  • Training hyperparameters = Adam lr=1e-5, decay factor 0.65, L1/L2 regularization, dropout 0.1 in CCR; epochs and batch size missing
    Reported in Section 4.1.4, but key settings like epoch count and batch size are missing.
  • Data split ratio = 80/10/10
    Random split of LiTS and 3DIRCADb used for training, validation, and test; not the official LiTS challenge protocol.
assumptions (5)
  • domain assumption Slice-wise 2D processing of 3D CT volumes preserves enough information for accurate liver and tumor segmentation
    Authors state they train and test on 2D images due to memory constraints and acknowledge inter-slice features are neglected (Discussion, Section 5).
  • domain assumption Random 80/10/10 splitting of LiTS and 3DIRCADb produces a test set comparable to published benchmark results
    No official challenge test split is used; the authors do not justify why numbers from different splits are comparable (Section 4.1.1, Table 2).
  • domain assumption The twelve baseline methods were evaluated under a protocol comparable to T-MPEDNet
    Section 4.1.2 lists baselines, but no training details, splits, or preprocessing are given for them.
  • ad hoc to paper Morphological boundary refinement, as an erosion-based post-processing step, can improve Dice against non-eroded ground-truth labels
    Section 3.2.3 describes only how to compute a boundary mask, not how it reshapes the predicted mask; the positive ablation in Table 4 relies on an unexplained mechanism.
  • domain assumption HU windowing, CLAHE, and z-score normalization preserve all clinically relevant liver and tumor information
    The reformation block in Section 3.2.1 applies these preprocessing steps without evaluating their individual effect on segmentation accuracy.

how reviews work

0 comments
Cite this review

Pith. "Pith review of T-MPEDNet: Unveiling the Synergy of Transformer-aware Multiscale Progressive Encoder-Decoder Network with Feature Recalibration for Tumor and Liver Segmentation." pith.science (2026). https://pith.science/paper/YJMCDU55

@misc{pith2026250719590,
  author       = {Pith},
  title        = {Pith review of: T-MPEDNet: Unveiling the Synergy of Transformer-aware Multiscale Progressive Encoder-Decoder Network with Feature Recalibration for Tumor and Liver Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YJMCDU55}},
  note         = {Machine review of arXiv:2507.19590}
}
read the original abstract

Precise and automated segmentation of the liver and its tumor within CT scans plays a pivotal role in swift diagnosis and the development of optimal treatment plans for individuals with liver diseases and malignancies. However, automated liver and tumor segmentation faces significant hurdles arising from the inherent heterogeneity of tumors and the diverse visual characteristics of livers across a broad spectrum of patients. Aiming to address these challenges, we present a novel Transformer-aware Multiscale Progressive Encoder-Decoder Network (T-MPEDNet) for automated segmentation of tumor and liver. T-MPEDNet leverages a deep adaptive features backbone through a progressive encoder-decoder structure, enhanced by skip connections for recalibrating channel-wise features while preserving spatial integrity. A Transformer-inspired dynamic attention mechanism captures long-range contextual relationships within the spatial domain, further enhanced by multi-scale feature utilization for refined local details, leading to accurate prediction. Morphological boundary refinement is then employed to address indistinct boundaries with neighboring organs, capturing finer details and yielding precise boundary labels. The efficacy of T-MPEDNet is comprehensively assessed on two widely utilized public benchmark datasets, LiTS and 3DIRCADb. Extensive quantitative and qualitative analyses demonstrate the superiority of T-MPEDNet compared to twelve state-of-the-art methods. On LiTS, T-MPEDNet achieves outstanding Dice Similarity Coefficients (DSC) of 97.6% and 89.1% for liver and tumor segmentation, respectively. Similar performance is observed on 3DIRCADb, with DSCs of 98.3% and 83.3% for liver and tumor segmentation, respectively. Our findings prove that T-MPEDNet is an efficacious and reliable framework for automated segmentation of the liver and its tumor in CT scans.

Figures

Figures reproduced from arXiv: 2507.19590 by the authors.

Figure 1
Figure 1. Diverse CT scans labeled (a) to (j) highlighting liver segmentation complexities, including (a) overlapping organs and tumors, (b) intricate liver anatomy, (c) the presence of artifacts and noise, (d) overlapping intensities, (e) irregular borders due to cirrhosis, (f) low contrast between the liver and surrounding healthy tissues, (g) fuzzy and indistinct tumor boundaries, (h) limited visibility of small tumors, (i… view at source ↗
Figure 2
Figure 2. The figure outlines the proposed framework for precisely segmenting liver and tumors. The process begins with a 3D liver voxel input, which transforms 2D slices, denoted as XCT , through the reformation block, resulting in enhanced overall quality. A multi-stage encoder extracts progressively deeper features, guided by Adaptive Feature Extraction (AdaFEx) modules. The Dynamic Contextual Attention (DCA) module synerg… view at source ↗
Figure 3
Figure 3. Visualization of the processing pipeline within the Reformation Block. 3.2.2. Progressive Feature Encoder Decoder [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Architecture of the Progressive Feature Encoder-Decoder (PED) block, a core component of the proposed T-MPEDNet framework. PED leverages five mirrored Adaptive Feature Extraction (AdaFEx) modules, a Compressive Channel Recalibration (CCR) module, a Dynamic Contex￾tual …
Figure 5
Figure 5. Figure 5: The Architecture of the Compressive Channel Recalibration module. For an input F⊕ = [f 1 ⊕ , f 2 ⊕ , . . . , f Cf ⊕ ], where f k ⊕ ∈ RH×W×Cf , we leverage global average pooling to reduce the spatial dimensions of each channel while preserving channel-specific informat…
Figure 6
Figure 6. Figure 6: The Architecture of Multi-Head Dynamic Contextual Attention module. The FC layer of the DCA module outputs F FC1 , a tensor of dimension Cf × d, where Cf signifies the number of convolution filters and d represents the length of the tensor. This key input, visualized i…
Figure 7
Figure 7. Figure 7: The Architecture of the Multi-Scale Atrous Spatial module. In [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Illustration of Morphological Boundary Refinement (MBR) module for refining the final boundary mask of liver and tumor boundary. The blue color represents the background pixel of the segmentation mask XMask, red represents the foreground pixels, and purple represents t…
Figure 9
Figure 9. Figure 9: Axial CT slices from the LiTS and 3DIRCADb datasets, highlighting the variety of liver and tumor anatomies present in the study. LiTS: The Liver Tumor Segmentation Challenge (LiTS) dataset [19] is a popular benchmark for evaluating auto￾mated liver and tumor segmentati…
Figure 10
Figure 10. Figure 10: Qualitative assessment of the visual quality of T-MPEDNet segmentation outputs on the LiTS dataset, compared to existing baseline methods. unable to identify tumor boundaries under the Encoder-Decoder-based segmentation technique. T-MPEDNet effi￾ciently utilizes the a…
Figure 12
Figure 12. Figure 12: visualizes the segmentation results on three CT scans using various configurations of the T-MPEDNet framework. The first row depicts the T-MPEDNet without additional modules. The significant deviations between the predicted segmentations (yellow boundary) and the grou…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 35 canonical work pages

  1. [1]

    J. Ma, Y . Zhang, S. Gu, C. Zhu, C. Ge, Y . Zhang, X. An, C. Wang, Q. Wang, X. Liu, S. Cao, Q. Zhang, S. Liu, Y . Wang, Y . Li, J. He, X. Yang, Abdomenct-1k: Is abdominal organ segmentation a solved problem, IEEE Transactions on Pattern Analysis and Machine Intelligence (2021). doi:10.1109/TPAMI.2021.3100536

    arXiv 2021

  2. [2]

    J. M. J. Valanarasu, V . A. Sindagi, I. Hacihaliloglu, V . M. Patel, Kiu-net: Overcomplete convolutional architectures for biomedical image and volumetric segmentation, IEEE Transactions on Medical Imaging 41 (2022). doi:10.1109/TMI.2021.3130469

    work page arXiv 2022

  3. [3]

    X. Li, H. Chen, X. Qi, Q. Dou, C. W. Fu, P. A. Heng, H-denseunet: Hybrid densely connected unet for liver and tumor segmentation from ct volumes, IEEE Transactions on Medical Imaging 37 (2018). doi:10.1109/TMI.2018.2845918

    work page arXiv 2018

  4. [4]

    Z. Yang, Y . qian Zhao, M. Liao, S. hu Di, Y . zhan Zeng, Semi-automatic liver tumor segmentation with adaptive region growing and graph cuts, Biomedical Signal Processing and Control 68 (2021). doi:10.1016/j.bspc.2021.102670

    work page arXiv 2021

  5. [5]

    W. He, C. Zhang, J. Dai, L. Liu, T. Wang, X. Liu, Y . Jiang, N. Li, J. Xiong, L. Wang, Y . Xie, X. Liang, A statistical deformation model- based data augmentation method for volumetric medical image segmentation, Medical Image Analysis 91 (2024) 102984. doi:10.1016/ j.media.2023.102984

    work page arXiv 2024

  6. [6]

    W. Liu, H. Yang, T. Tian, Z. Cao, X. Pan, W. Xu, Y . Jin, F. Gao, Full-resolution network and dual-threshold iteration for retinal vessel and coronary angiograph segmentation, IEEE Journal of Biomedical and Health Informatics 26 (2022). doi:10.1109/JBHI.2022.3188710

    work page arXiv 2022

  7. [7]

    A. J. Dickson, J. A. Linsely, V . A. A. Daniel, K. Rahul, Sparse deep belief network coupled with extended local fuzzy active contour model- based liver cancer segmentation from abdomen ct images, Medical & Biological Engineering & Computing (2024) 1–14 doi:10.1007/ s11517-023-03001-y

    work page 2024

  8. [8]

    S. Di, Y . Q. Zhao, M. Liao, F. Zhang, X. Li, Td-net: A hybrid end-to-end network for automatic liver tumor segmentation from ct images, IEEE Journal of Biomedical and Health Informatics 27 (2023). doi:10.1109/JBHI.2022.3181974

    work page arXiv 2023

Show all 46 references
  1. [9]

    Ronneberger, P

    O. Ronneberger, P. Fischer, T. Brox, U-net: Convolutional networks for biomedical image segmentation, in: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), V ol. 9351, 2015, pp. 234–241. doi:10...

    2015 doi

  2. [10]

    M. Liao, H. Tang, X. Li, P. Vijayakumar, V . Arya, B. B. Gupta, A lightweight network for abdominal multi-organ segmentation based on multi-scale context fusion and dual self-attention, Information Fusion 108 (2024) 102401. doi:10.1016/j.inffus.2024.102401

    2024

  3. [11]

    D. Y . Tu, P. C. Lin, H. H. Chou, M. R. Shen, S. Y . Hsieh, Slice-fusion: Reducing false positives in liver tumor detection for mask r-cnn, IEEE/ACM Transactions on Computational Biology and Bioinformatics 20 (2023). doi:10.1109/TCBB.2023.3265394

    2023

  4. [12]

    Y . Li, B. Zou, P. Dai, M. Liao, H. X. Bai, Z. Jiao, Ac-e network: Attentive context-enhanced network for liver segmentation, IEEE Journal of Biomedical and Health Informatics 27 (2023). doi:10.1109/JBHI.2023.3278079

    2023

  5. [13]

    D. T. Kushnure, S. N. Talbar, Hfru-net: High-level feature fusion and recalibration unet for automatic liver and tumor segmentation in ct images, Computer Methods and Programs in Biomedicine 213 (2022). doi:10.1016/j.cmpb.2021.106501

    2022

  6. [14]

    Jiang, J

    L. Jiang, J. Ou, R. Liu, Y . Zou, T. Xie, H. Xiao, T. Bai, Rmau-net: Residual multi-scale attention u-net for liver and tumor segmentation in ct images, Computers in Biology and Medicine 158 (2023). doi:10.1016/j.compbiomed.2023.106838

    2023

  7. [15]

    Y . Liu, H. Wang, Z. Chen, K. Huangliang, H. Zhang, Transunet +: Redesigning the skip connection to enhance features in medical image segmentation, Knowledge-Based Systems 256 (2022). doi:10.1016/j.knosys.2022.109859

    2022

  8. [16]

    Y . Xie, J. Zhang, C. Shen, Y . Xia, Cotr: E fficiently bridging cnn and transformer for 3d medical image segmentation, in: Medical Image Computing and Computer Assisted Intervention–MICCAI 2021: 24th International Conference, Strasbourg, France, September 27–October 1, 2021, ...

    2021 doi

  9. [17]

    Vaswani, N

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

    2017

  10. [18]

    Zhang, S

    X. Zhang, S. Yang, Y . Jiang, Y . Chen, F. Sun, Fafs-unet: Redesigning skip connections in unet with feature aggregation and feature selection, Computers in Biology and Medicine 170 (2024) 108009. doi:10.1016/j.compbiomed.2024.108009

    2024

  11. [19]

    Bilic, P

    P. Bilic, P. Christ, H. B. Li, E. V orontsov, A. Ben-Cohen, G. Kaissis, A. Szeskin, C. Jacobs, G. E. H. Mamani, G. Chartrand, F. Loh¨ofer, J. W. Holch, W. Sommer, F. Hofmann, A. Hostettler, N. Lev-Cohain, M. Drozdzal, M. M. Amitai, R. Vivanti, J. Sosna, I. Ezhov, A. Sekuboyina...

    2023

  12. [20]

    Soler, A

    L. Soler, A. Hostettler, V . Agnus, A. Charnoz, J.-B. Fasquel, J. Moreau, A.-B. Osswald, M. Bouhadjar, J. Marescaux, 3d image reconstruction for comparison of algorithm database, URL: https://www. ircad. fr/research/data-sets/liver-segmentation-3d-ircadb-01 (2010)

    2010

  13. [21]

    Gopinath, V

    B. Gopinath, V . V . Kumar, H. Kag, K. Joshi, A. Rajput, D. Kapila, Medical image segmentation using deep learning techniques: Advancement and challenges, in: 2023 3rd International Conference on Advance Computing and Innovative Technologies in Engineering, ICACITE 2023, 2023,...

    2023

  14. [22]

    Mahendran, P

    N. Mahendran, P. Muthuvel, T. Arunprasath, M. P. Rajasekaran, J. B. Nirmala, R. Kottaimalai, Identification and segmentation of tumour in brain mri using deep learning techniques, in: Proceedings of the 2023 2nd International Conference on Electronics and Renewable Systems, IC...

    2023

  15. [23]

    T. A. Soomro, L. Zheng, A. J. Afifi, A. Ali, S. Soomro, M. Yin, J. Gao, Image segmentation for mr brain tumor detection using machine learning: A review (2023). doi:10.1109/RBME.2022.3185292

    2023

  16. [24]

    Qayyum, I

    A. Qayyum, I. Razzak, M. Mazher, X. Lu, S. A. Niederer, Unsupervised unpaired multiple fusion adaptation aided with self-attention 22 generative adversarial network for scar tissues segmentation framework, Information Fusion 106 (2024). doi:10.1016/j.inffus.2024. 102226

    2024 doi

  17. [25]

    S. Di, Y . Zhao, M. Liao, Z. Yang, Y . Zeng, Automatic liver tumor segmentation from ct images using hierarchical iterative superpixels and local statistical features, Expert Systems with Applications 203 (2022). doi:10.1016/j.eswa.2022.117347

    2022

  18. [26]

    Y . Chen, C. Zheng, T. Zhou, L. Feng, L. Liu, Q. Zeng, G. Wang, A deep residual attention-based u-net with a biplane joint method for liver segmentation from ct scans, Computers in Biology and Medicine 152 (2023). doi:10.1016/j.compbiomed.2022.106421

    2023

  19. [27]

    G. Chen, Z. Li, J. Wang, J. Wang, S. Du, J. Zhou, J. Shi, Y . Zhou, An improved 3d kiu-net for segmentation of liver tumor, Computers in Biology and Medicine 160 (2023). doi:10.1016/j.compbiomed.2023.107006

    2023

  20. [28]

    S. LI, G. K. TSO, K. HE, Bottleneck feature supervised u-net for pixel-wise liver and tumor segmentation, Expert Systems with Applications 145 (2020). doi:10.1016/j.eswa.2019.113131

    2020

  21. [29]

    J. Zhao, X. Hou, M. Pan, H. Zhang, Attention-based generative adversarial network in medical imaging: A narrative review, Computers in Biology and Medicine 149 (2022). doi:10.1016/j.compbiomed.2022.105948

    2022

  22. [30]

    X. Li, S. Lv, M. Li, J. Zhang, Y . Jiang, Y . Qin, H. Luo, S. Yin, Sdmt: Spatial dependence multi-task transformer network for 3d knee mri segmentation and landmark localization, IEEE Transactions on Medical Imaging 42 (2023). doi:10.1109/TMI.2023.3247543

    2023

  23. [31]

    N. K. Tomar, D. Jha, M. A. Riegler, H. D. Johansen, D. Johansen, J. Rittscher, P. Halvorsen, S. Ali, Fanet: A feedback attention network for improved biomedical image segmentation, IEEE Transactions on Neural Networks and Learning Systems 34 (2023).doi:10.1109/TNNLS. 2022.3159394

    2023

  24. [32]

    Y . Wu, Q. Kong, L. Zhang, A. Castiglione, M. Nappi, S. Wan, Cdt-cad: Context-aware deformable transformers for end-to-end chest abnor- mality detection on x-ray images, IEEE /ACM Transactions on Computational Biology and Bioinformatics (2023). doi:10.1109/TCBB. 2023.3258455

    2023

  25. [33]

    J. Li, G. Huang, J. He, Z. Chen, C. M. Pun, Z. Yu, W. K. Ling, L. Liu, J. Zhou, J. Huang, Shift-channel attention and weighted-region loss function for liver and dense tumor segmentation, Medical Physics 49 (2022). doi:10.1002/mp.15816

    2022 doi

  26. [34]

    H. Liu, Y . Fu, S. Zhang, J. Liu, Y . Wang, G. Wang, J. Fang, Gcha-net: Global context and hybrid attention network for automatic liver segmentation, Computers in Biology and Medicine 152 (2023). doi:10.1016/j.compbiomed.2022.106352

    2023

  27. [35]

    M. S. Alam, D. Wang, Q. Liao, A. Sowmya, A multi-scale context aware attention model for medical image segmentation, IEEE Journal of Biomedical and Health Informatics 27 (2023). doi:10.1109/JBHI.2022.3227540

    2023

  28. [36]

    B. Cui, W. Jing, L. Huang, Z. Li, Y . Lu, Sanet: A sea-land segmentation network via adaptive multiscale feature learning, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14 (2021). doi:10.1109/JSTARS.2020.3040176

    2021

  29. [37]

    X. Du, J. Wang, W. Sun, Densely connected u-net retinal vessel segmentation algorithm based on multi-scale feature convolution extraction, Medical Physics 48 (2021). doi:10.1002/mp.14944

    2021 doi

  30. [38]

    Abdar, M

    M. Abdar, M. A. Fahami, L. Rundo, P. Radeva, A. F. Frangi, U. R. Acharya, A. Khosravi, H. K. Lam, A. Jung, S. Nahavandi, Hercules: Deep hierarchical attentive multilevel fusion model with uncertainty quantification for medical image classification, IEEE Transactions on Industr...

    2023

  31. [39]

    G. Yue, S. Li, R. Cong, T. Zhou, B. Lei, T. Wang, Attention-guided pyramid context network for polyp segmentation in colonoscopy images, IEEE Transactions on Instrumentation and Measurement 72 (2023). doi:10.1109/TIM.2023.3244219

    2023

  32. [40]

    S. S. Dar, B. Kaurav, A. Jain, C. S. Raghaw, M. Z. U. Rehman, N. Kumar, An explainable deep neural network with frequency-aware channel and spatial refinement for flood prediction in sustainable cities, Sustainable Cities and Society 130 (2025) 106480. doi:10.1016/j.scs. 2025.106480

    2025

  33. [41]

    T. Lei, R. Wang, Y . Zhang, Y . Wan, C. Liu, A. K. Nandi, Defed-net: Deformable encoder-decoder network for liver and liver tumor segmen- tation, IEEE Transactions on Radiation and Plasma Medical Sciences 6 (2022). doi:10.1109/TRPMS.2021.3059780

    2022

  34. [42]

    Y . Chen, C. Zheng, F. Hu, T. Zhou, L. Feng, G. Xu, Z. Yi, X. Zhang, Efficient two-step liver and tumour segmentation on abdominal ct via deep learning and a conditional random field, Computers in Biology and Medicine 150 (2022). doi:10.1016/j.compbiomed.2022.106076

    2022

  35. [43]

    Y . Chen, C. Zheng, W. Zhang, H. Lin, W. Chen, G. Zhang, G. Xu, F. Wu, Ms-fanet: Multi-scale feature attention network for liver tumor segmentation, Computers in Biology and Medicine 163 (2023). doi:10.1016/j.compbiomed.2023.107208

    2023

  36. [44]

    P. Lv, J. Wang, X. Zhang, C. Shi, Deep supervision and atrous inception-based u-net combining crf for automatic liver segmentation from ct, Scientific Reports 12 (2022). doi:10.1038/s41598-022-21562-0

    2022 doi

  37. [45]

    X. Guo, Z. Wang, P. Wu, Y . Li, F. E. Alsaadi, N. Zeng, Elts-net: An enhanced liver tumor segmentation network with augmented receptive field and global contextual information, Computers in Biology and Medicine 169 (2024) 107879.doi:10.1016/j.compbiomed.2023.107879

    2024

  38. [46]

    Z. Wang, J. Zhu, S. Fu, Y . Ye, Context fusion network with multi-scale-aware skip connection and twin-split attention for liver tumor segmentation, Medical and Biological Engineering and Computing 61 (2023). doi:10.1007/s11517-023-02876-1 . 23

    2023 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.