pith. sign in

arxiv: 2606.10378 · v1 · pith:KX5BVZ5Hnew · submitted 2026-06-09 · 💻 cs.CV

FSS-Net: Frequency-Spatial Synergy Network with Wavelet Attention for Carotid Artery Ultrasound Segmentation

Jiawei Liu , Zhijiang Wan , Junhua Hu , Rongli Zhang , Zhongbiao Xu , Yankun Cao , Yuan Chen , Jin Hong This is my paper

Pith reviewed 2026-06-27 13:58 UTC · model grok-4.3

classification 💻 cs.CV
keywords carotid artery segmentationultrasound imagingwavelet attentionspeckle noisemedical image segmentationedge fusionfrequency-spatial synergy
0
0 comments X

The pith

FSS-Net segments carotid arteries in ultrasound at 96.46 percent Dice score by merging wavelet frequency attention with spatial processing.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces FSS-Net, an encoder-decoder network that applies wavelet transforms inside attention and bottleneck modules to suppress speckle noise while preserving boundary details in low-contrast carotid ultrasound scans. Three specialized components handle multi-domain feature purification, long-range dependency capture, and adaptive edge fusion without requiring post-processing. Experiments on carotid datasets show the network maintains high accuracy under reduced signal-to-noise ratios and extends to breast-cancer ultrasound segmentation.

Core claim

FSS-Net integrates a Channel-Spatial-Wavelet Attention module to purify semantic features in the frequency domain, a Wavelet-Enhanced Bottleneck module to capture global dependencies, and a Laplacian-Guided Adaptive Edge Fusion module to restore high-frequency boundary continuity, yielding a Dice score of 96.46 percent and improved robustness on carotid ultrasound data compared with prior methods.

What carries the argument

The Frequency-Spatial Synergy Network (FSS-Net) architecture whose CSWA, WEB, and LAEF modules embed wavelet transforms to jointly process frequency and spatial information for noise-robust segmentation.

If this is right

  • Accurate plaque identification supports stroke risk assessment from routine ultrasound exams.
  • Robustness under low SNR reduces the need for high-quality image acquisition protocols.
  • Extension to breast-cancer segmentation indicates applicability to other ultrasound tasks involving abnormal tissue masses.
  • Outperformance of existing methods suggests the frequency-spatial design can replace separate denoising and edge-detection stages.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The wavelet attention approach may transfer to other speckle-noise imaging modalities such as intravascular ultrasound or echocardiography.
  • Because the modules operate inside a standard encoder-decoder, the network could be inserted into existing clinical pipelines with minimal architectural change.
  • Real-time inference speed on modest hardware would determine whether the method supports bedside plaque monitoring.

Load-bearing premise

The three modules deliver noise suppression and boundary accuracy gains on representative clinical ultrasound images without dataset-specific tuning.

What would settle it

Measure Dice score on a held-out carotid ultrasound collection acquired with different scanner hardware or patient population; if performance falls below 94 percent without retraining, the claimed general robustness does not hold.

read the original abstract

Accurate segmentation of carotid arteries in ultrasound imaging is critical for stroke risk assessment. However, speckle noise, low contrast, and blurred boundaries remain major challenges. In this paper, we propose a Frequency-Spatial Synergy Network (FSS-Net) to achieve noise-robust and high-precision carotid artery segmentation. The network integrates wavelet transform, multi-domain attention, and edge enhancement into a unified encoder-decoder architecture. Specifically, a Channel-Spatial-Wavelet Attention (CSWA) module is designed to suppress noise and purify semantic features in the frequency domain. A Wavelet-Enhanced Bottleneck (WEB) module is introduced to capture long-range global dependencies efficiently. Furthermore, a Laplacian-Guided Adaptive Edge Fusion (LAEF) module compensates high-frequency details and maintains boundary continuity. Extensive experiments on carotid ultrasound datasets show that FSS-Net achieves a Dice score (DSC) of 96.46% and strong robustness under low SNR conditions, outperforming several state-of-the-art methods. This method realizes accurate segmentation of carotid artery in ultrasonic imaging, effectively identifies carotid atherosclerotic plaque, and is verified by other task (such as segmentation of breast cancer), suggesting that it has good clinical application potential in identifying abnormal tissue masses in ultrasonic images.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper proposes FSS-Net, an encoder-decoder network for carotid artery segmentation in ultrasound images that integrates wavelet transforms with a Channel-Spatial-Wavelet Attention (CSWA) module for noise suppression, a Wavelet-Enhanced Bottleneck (WEB) module for global dependencies, and a Laplacian-Guided Adaptive Edge Fusion (LAEF) module for boundary preservation. It claims a Dice score of 96.46% on carotid ultrasound datasets, superior performance over state-of-the-art methods, and robustness under low SNR conditions, with suggested extension to other ultrasound tasks such as breast cancer segmentation.

Significance. If the reported performance and robustness hold under proper validation, the frequency-spatial synergy approach could meaningfully improve automated plaque and vessel analysis in clinical ultrasound, aiding stroke risk assessment; the modular design offers a concrete template for handling speckle noise that might transfer to other low-contrast imaging domains.

major comments (2)
  1. [Abstract] Abstract: The central performance claim (DSC = 96.46%, robustness under low SNR, outperformance of SOTA) is stated without any reference to dataset names/sizes, train/test splits, baseline implementations, statistical tests, or error bars; this absence prevents verification that the data support the claim and is load-bearing for the paper's primary contribution.
  2. [Abstract] Abstract: The assertion that the CSWA, WEB, and LAEF modules deliver the claimed noise suppression and boundary improvements is presented without ablation results, quantitative module-wise contributions, or controls for dataset-specific tuning, leaving the weakest assumption untested in the provided text.
minor comments (1)
  1. [Abstract] Abstract, final sentence: phrasing is awkward and contains a grammatical error ("verified by other task (such as segmentation of breast cancer)"); reword for clarity and precision.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on the abstract. We address each point below and will revise the manuscript to improve the verifiability of the claims while respecting abstract length constraints.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The central performance claim (DSC = 96.46%, robustness under low SNR, outperformance of SOTA) is stated without any reference to dataset names/sizes, train/test splits, baseline implementations, statistical tests, or error bars; this absence prevents verification that the data support the claim and is load-bearing for the paper's primary contribution.

    Authors: The full manuscript details the carotid ultrasound datasets (including names, sizes, and train/test splits), baseline implementations, error bars, and statistical tests in the Experiments section. We agree the abstract would benefit from greater specificity for immediate verifiability. We will revise the abstract to incorporate brief references to the datasets and validation protocol (e.g., cross-validation and SNR conditions) without exceeding typical length limits. revision: yes

  2. Referee: [Abstract] Abstract: The assertion that the CSWA, WEB, and LAEF modules deliver the claimed noise suppression and boundary improvements is presented without ablation results, quantitative module-wise contributions, or controls for dataset-specific tuning, leaving the weakest assumption untested in the provided text.

    Authors: The manuscript contains ablation studies in the Experiments section that quantify the individual and synergistic contributions of the CSWA, WEB, and LAEF modules to noise suppression and boundary accuracy, including performance under varying SNR conditions. We acknowledge these details are absent from the abstract summary. We will revise the abstract to explicitly reference the module contributions supported by the ablations. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper proposes an encoder-decoder network (FSS-Net) with three described modules (CSWA, WEB, LAEF) whose design is presented as an engineering choice for handling ultrasound noise and boundaries. Performance claims (DSC 96.46%, robustness under low SNR) are stated as outcomes of experiments on carotid ultrasound datasets with comparisons to other methods. No equations, first-principles derivations, or predictions appear that reduce by construction to fitted parameters, self-definitions, or self-citation chains. The central claims rest on empirical validation rather than any load-bearing step that collapses to its own inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The abstract describes high-level network modules and empirical results but introduces no mathematical axioms, fitted free parameters, or new postulated entities.

pith-pipeline@v0.9.1-grok · 5778 in / 1022 out tokens · 24916 ms · 2026-06-27T13:58:43.717772+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

52 extracted references · 1 linked inside Pith

  1. [1]

    Evaluation of lipoprotein-associated phospholipase A2, serum amyloid A, and fibrinogen as diagnostic biomarkers for patients with acute cerebral infarction,

    T. Liang et al., "Evaluation of lipoprotein-associated phospholipase A2, serum amyloid A, and fibrinogen as diagnostic biomarkers for patients with acute cerebral infarction," (in English), Journal of Clinical Laboratory Analysis, Article vol. 34, no. 3, p. 7, Mar 2020, Art. no. e23084

    2020

  2. [2]

    World Stroke Organization (WSO): Global Stroke Fact Sheet 2022 (vol 17, pg 18, 2022),

    V . L. Feigin et al., "World Stroke Organization (WSO): Global Stroke Fact Sheet 2022 (vol 17, pg 18, 2022)," (in English), International Journal of Stroke, Correction vol. 17, no. 4, pp. 478 -478, Apr 2022, Art. no. 17474930221080343

    2022

  3. [3]

    Global Burden of Cardiovascular Diseases and Risk Factors, 1990-2019 Update From the GBD 2019 Study,

    G. A. Roth et al., "Global Burden of Cardiovascular Diseases and Risk Factors, 1990-2019 Update From the GBD 2019 Study," (in English), Journal of the American College of Cardiology, Review vol. 76, no. 25, pp. 2982 -3021, Dec 22 2020

    1990

  4. [4]

    A deep learning -based system to identify originating mural layer of upper gastrointestinal submucosal tumors under EUS,

    X. Li et al., "A deep learning -based system to identify originating mural layer of upper gastrointestinal submucosal tumors under EUS," ENDOSCOPIC ULTRASOUND, vol. 12, no. 6, pp. 465-471, NOV-DEC 2023

    2023

  5. [5]

    Contrast harmonic endoscopic ultrasound: Instrumentation, echoprocessors, and echoendoscopes,

    J. Iglesias -Garcia et al. , "Contrast harmonic endoscopic ultrasound: Instrumentation, echoprocessors, and echoendoscopes," ENDOSCOPIC ULTRASOUND, vol. 6, no. 1, pp. 37-42, JAN-FEB 2017

    2017

  6. [6]

    An automated extraction of spectral-temporal and spatial-temporal features of EEG for emotion detection,

    M. Islam and T. Lee, "An automated extraction of spectral-temporal and spatial-temporal features of EEG for emotion detection," (in eng), no. 2198-4018 (Print)

  7. [7]

    Ensemble Machine Learning and Its V alidation for Prediction of Coronary Artery Disease and Acute Coronary Syndrome Using Focused Carotid Ultrasound,

    A. D. Jamthikar et al., " Ensemble Machine Learning and Its V alidation for Prediction of Coronary Artery Disease and Acute Coronary Syndrome Using Focused Carotid Ultrasound," IEEE Transactions on Instrumentation and Measurement, vol. 71, p. 10, 2022

    2022

  8. [8]

    ERANet: Edge replacement augmentation for semi -supervised meniscus segmentation with prototype consistency alignment and conditional self-training,

    S. Li et al., "ERANet: Edge replacement augmentation for semi -supervised meniscus segmentation with prototype consistency alignment and conditional self-training," Neural Networks, vol. 196, p. 108337, 2026/04/01/ 2026

    2026

  9. [9]

    DGSSA: Domain generalization with structural and stylistic augmentation for retinal vessel segmentation,

    B. Liu et al. , "DGSSA: Domain generalization with structural and stylistic augmentation for retinal vessel segmentation," Neural Networks, vol. 194, p. 108118, 2026/02/01/ 2026

    2026

  10. [10]

    MGML: A plug-and-play meta-guided multi-modal learning framework for incomplete multimodal brain tumor segmentation,

    Y . L. Zou et al., "MGML: A plug-and-play meta-guided multi-modal learning framework for incomplete multimodal brain tumor segmentation," NEURAL NETWORKS, vol. 197, MAY 2026, Art. no. 108525

    2026

  11. [11]

    Carotid wall segmentation in longitudinal ultrasound images using structured random forest,

    Y . Nagaraj et al., "Carotid wall segmentation in longitudinal ultrasound images using structured random forest," (in English), Computers & Electrical Engineering, Article vol. 69, pp. 753-767, Jul 2018

    2018

  12. [12]

    A multimodal artificial intelligence system for the detection and diagnosis of solid pancreatic lesions under EUS,

    C. X. Zhang et al., "A multimodal artificial intelligence system for the detection and diagnosis of solid pancreatic lesions under EUS," ENDOSCOPIC ULTRASOUND, vol. 14, no. 5, pp. 274-281, SEP-OCT 2025

    2025

  13. [13]

    The role of artificial intelligence and deep learning in determining the histopathological grade of pancreatic neuroendocrine tumors by using EUS images,

    S. Kiremitci et al., "The role of artificial intelligence and deep learning in determining the histopathological grade of pancreatic neuroendocrine tumors by using EUS images," ENDOSCOPIC ULTRASOUND, vol. 14, no. 2, pp. 48 -56, MAR-APR 2025

    2025

  14. [14]

    Risk stratification system of gastrointestinal stromal tumors under EUS elastography based on artificial intelligence,

    C. X. Zhang et al., "Risk stratification system of gastrointestinal stromal tumors under EUS elastography based on artificial intelligence," ENDOSCOPIC ULTRASOUND, vol. 14, no. 4, pp. 220-227, JUL-AUG 2025. 27

    2025

  15. [15]

    Segmentation of carotid artery in ultrasound images: Method development and evaluation technique,

    F. Mao et al., "Segmentation of carotid artery in ultrasound images: Method development and evaluation technique," (in English), Medical Physics, Article vol. 27, no. 8, pp. 1961-1970, Aug 2000

    1961

  16. [16]

    Using the hough transform to segment ultrasound images of longitudinal and transverse sections of the carotid artery,

    S. Golemati et al., "Using the hough transform to segment ultrasound images of longitudinal and transverse sections of the carotid artery," (in English), Ultrasound in Medicine and Biology, Article vol. 33, no. 12, pp. 1918-1932, Dec 2007

    1918

  17. [17]

    Automatic carotid ultrasound segmentation using deep convolutional neural networks and phase congruency maps,

    C. Azzopardi et al., "Automatic carotid ultrasound segmentation using deep convolutional neural networks and phase congruency maps," in IEEE 14th International Symposium on Biomedical Imaging (ISBI) - From Nano to Macro , Melbourne, AUSTRALIA, 2017, pp. 624-628, NEW YORK: Ieee, 2017

    2017

  18. [18]

    Localization of common carotid artery transverse section in B -mode ultrasound images using faster RCNN: a deep learning approach,

    P. K. Jain et al., "Localization of common carotid artery transverse section in B -mode ultrasound images using faster RCNN: a deep learning approach," (in English), Medical & Biological Engineering & Computing, Article vol. 58, no. 3, pp. 471-482, Mar 2020

    2020

  19. [19]

    Bi-attention DoubleUNet: A deep learning approach for carotid artery segmentation in transverse view images for non-invasive stenosis diagnosis,

    N. Ottakath et al., "Bi-attention DoubleUNet: A deep learning approach for carotid artery segmentation in transverse view images for non-invasive stenosis diagnosis," Biomedical Signal Processing and Control, vol. 94, 2024

    2024

  20. [20]

    A Generalized Approach for Automatic 3-D Geometry Assessment of Blood V essels in Transverse Ultrasound Images Using Convolutional Neural Networks,

    J. de Ruijter et al., "A Generalized Approach for Automatic 3-D Geometry Assessment of Blood V essels in Transverse Ultrasound Images Using Convolutional Neural Networks," IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 68, no. 11, pp. 3326-3335, 2021

    2021

  21. [21]

    EANet: Iterative edge attention network for medical image segmentation,

    K. Wang et al., "EANet: Iterative edge attention network for medical image segmentation," Pattern Recognition, vol. 127, 2022

    2022

  22. [22]

    Toward automated right ventricle segmentation via edge feature-induced self-attention multiscale feature aggregation full convolution network,

    J. Liu et al., "Toward automated right ventricle segmentation via edge feature-induced self-attention multiscale feature aggregation full convolution network," IEEE Transactions on Instrumentation and Measurement, vol. 72, no. -, p. 12, 2023

    2023

  23. [23]

    Attentive Boundary -Aware Fusion for Defect Semantic Segmentation Using Transformer,

    C. C. Yeung and K. M. Lam, " Attentive Boundary -Aware Fusion for Defect Semantic Segmentation Using Transformer," IEEE Transactions on Instrumentation and Measurement, vol. 72, p. 13, 2023

    2023

  24. [24]

    Attention Mechanism Enhanced Multi -layer Edge Perception Network for Deep Semantic Medical Segmentation,

    M. Sun et al., "Attention Mechanism Enhanced Multi -layer Edge Perception Network for Deep Semantic Medical Segmentation," Cognitive Computation, vol. 15, no. 1, pp. 348-358, 2023

    2023

  25. [25]

    BAM: Bottleneck Attention Module,

    S. W. J. Park, J.-Y . Lee, and I. S. Kweon, "BAM: Bottleneck Attention Module," in British Machine Vision Conference (BMVC), 2018

    2018

  26. [26]

    CBAM: Convolutional Block Attention Module,

    S. H. Woo et al., "CBAM: Convolutional Block Attention Module," in 15th European Conference on Computer Vision (ECCV), Munich, GERMANY , 2018, vol. 11211, pp. 3-19, CHAM: Springer International Publishing Ag, 2018

    2018

  27. [27]

    Wavelet-Attention CNN for image classification,

    X. Y . Zhao et al., "Wavelet-Attention CNN for image classification," (in English), Multimedia Systems, Article vol. 28, no. 3, pp. 915-924, Jun 2022

    2022

  28. [28]

    WaveNet-SF: A hybrid network for retinal disease detection based on wavelet transform in spatial - frequency domain,

    J. Cheng et al., "WaveNet-SF: A hybrid network for retinal disease detection based on wavelet transform in spatial - frequency domain," Neural Netw, vol. 194, p. 108189, Oct 3 2025

    2025

  29. [29]

    MSLI -Net: retinal disease detection network based on multi -segment localization and multi -scale interaction,

    Z. Qi et al. , "MSLI -Net: retinal disease detection network based on multi -segment localization and multi -scale interaction," Front Cell Dev Biol, vol. 13, p. 1608325, 2025

    2025

  30. [30]

    Multi-stage image denoising with the wavelet transform,

    C. Tian et al., "Multi-stage image denoising with the wavelet transform," Pattern Recognition, vol. 134, 2023

    2023

  31. [31]

    Haar wavelet downsampling: A simple but effective downsampling module for semantic segmentation,

    G. Xu et al., "Haar wavelet downsampling: A simple but effective downsampling module for semantic segmentation," Pattern Recognition, vol. 143, 2023

    2023

  32. [32]

    Wave-ViT: Unifying Wavelet and Transformers for Visual Representation Learning,

    T. Yao et al., "Wave-ViT: Unifying Wavelet and Transformers for Visual Representation Learning," vol. abs/2207.04978, 2022

    arXiv 2022

  33. [33]

    Haar wavelet downsampling: A simple but effective downsampling module for semantic segmentation,

    G. Xu et al., "Haar wavelet downsampling: A simple but effective downsampling module for semantic segmentation," Pattern Recognition, vol. 143, p. 109819, 2023/11/01/ 2023

    2023

  34. [34]

    WRANet: wavelet integrated residual attention U -Net network for medical image segmentation,

    Y . Zhao et al. , "WRANet: wavelet integrated residual attention U -Net network for medical image segmentation," Complex & Intelligent Systems, vol. 9, no. 6, pp. 6971-6983, 2023/12/01 2023

    2023

  35. [35]

    MI-SegNet: Mutual Information-Based US Segmentation for Unseen Domain Generalization,

    Y . Bi et al., "MI-SegNet: Mutual Information-Based US Segmentation for Unseen Domain Generalization," in Medical 28 Image Computing and Computer Assisted Intervention – MICCAI 2023, Cham, 2023, pp. 130 -140: Springer Nature Switzerland

    2023

  36. [36]

    U-Net: Convolutional Networks for Biomedical Image Segmentation,

    O. Ronneberger et al., "U-Net: Convolutional Networks for Biomedical Image Segmentation," in 18th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), Munich, GERMANY , 2015, vol. 9351, pp. 234-241, CHAM: Springer International Publishing Ag, 2015

    2015

  37. [37]

    Selective Kernel Networks,

    X. Li et al., "Selective Kernel Networks," presented at the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019

    2019

  38. [38]

    Selective kernel convolution deep residual network based on channel-spatial attention mechanism and feature fusion for mechanical fault diagnosis,

    S. Zhang et al., "Selective kernel convolution deep residual network based on channel-spatial attention mechanism and feature fusion for mechanical fault diagnosis," ISA Transactions, vol. 133, pp. 369-383, 2023

    2023

  39. [39]

    Common Carotid Artery Ultrasound Images,

    A. Momot, "Common Carotid Artery Ultrasound Images," M. Data, Ed., ed, 2022

    2022

  40. [40]

    Dataset of breast ultrasound images,

    W. G. Al-Dhabyani, Mohammed Khaled, H. Fahmy, Aly, "Dataset of breast ultrasound images," Data in Brief, vol. 28, p. 104863, 2020

    2020

  41. [41]

    Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation,

    L. C. E. Chen et al., "Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation," in 15th European Conference on Computer Vision (ECCV) , Munich, GERMANY , 2018, vol. 11211, pp. 833 -851, CHAM: Springer International Publishing Ag, 2018

    2018

  42. [42]

    TransUNet: Rethinking the U-Net architecture design for medical image segmentation through the lens of transformers,

    J. N. Chen et al., "TransUNet: Rethinking the U-Net architecture design for medical image segmentation through the lens of transformers," (in English), Medical Image Analysis, Article vol. 97, p. 10, Oct 2024, Art. no. 103280

    2024

  43. [43]

    MultiResUNet : Rethinking the U -Net architecture for multimodal biomedical image segmentation,

    N. Ibtehaz and M. S. Rahman, "MultiResUNet : Rethinking the U -Net architecture for multimodal biomedical image segmentation," Neural Netw, vol. 121, pp. 74-87, Jan 2020

    2020

  44. [44]

    Rethinking Decoder Design: Improving Biomarker Segmentation Using Depth-to-Space Restoration and Residual Linear Attention,

    S. K. Wazir, D., "Rethinking Decoder Design: Improving Biomarker Segmentation Using Depth-to-Space Restoration and Residual Linear Attention," presented at the Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2025, 2025. Available: https://github.com/saadwazir/MCADS-Decoder

    2025

  45. [45]

    CondSeg: Ellipse Estimation of Pupil and Iris via Conditioned Segmentation,

    Z. Jia et al. , "CondSeg: Ellipse Estimation of Pupil and Iris via Conditioned Segmentation," in 18th European Conference on Computer Vision (ECCV), Milan, ITALY , 2024, vol. 15645, pp. 185-198, 2025

    2024

  46. [46]

    MHAHF -UNet: a multi -scale hybrid attention hierarchy fusion network for carotid artery segmentation,

    C. Jiang et al. , "MHAHF -UNet: a multi -scale hybrid attention hierarchy fusion network for carotid artery segmentation," Int J Comput Assist Radiol Surg, Jun 17 2025

    2025

  47. [47]

    FRDD-Net: Automated carotid plaque ultrasound images segmentation using feature remapping and dense decoding,

    Y . Li et al., "FRDD-Net: Automated carotid plaque ultrasound images segmentation using feature remapping and dense decoding," Sensors (Basel), vol. 22, no. 3, Jan 24 2022

    2022

  48. [48]

    nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation,

    F. Isensee et al., "nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation," Nature Methods, vol. 18, no. 2, p. 211, 2021

    2021

  49. [49]

    Attention U -Net: Learning Where to Look for the Pancreas,

    O. S. Oktay, J.; Le Folgoc, L.; Lee, M.; Heinrich, M.; Misawa, K.; Mori, K.; McDonagh, S.; Hammerla, N. Y .; Kainz, B.; Glocker, B.; Rueckert, D., "Attention U -Net: Learning Where to Look for the Pancreas," arXiv preprint arXiv:1804.03999, 2018

    Pith/arXiv arXiv 2018

  50. [50]

    Road Extraction by Deep Residual U -Net,

    Z. X. Zhang et al., "Road Extraction by Deep Residual U -Net," (in English), Ieee Geoscience and Remote Sensing Letters, Article vol. 15, no. 5, pp. 749-753, May 2018

    2018

  51. [51]

    HCTNet: A Hybrid ConvNet-Transformer Network for Retinal Optical Coherence Tomography Image Classification,

    Z. Q. Ma et al., "HCTNet: A Hybrid ConvNet-Transformer Network for Retinal Optical Coherence Tomography Image Classification," (in English), Biosensors-Basel, Article vol. 12, no. 7, p. 15, Jul 2022, Art. no. 542

    2022

  52. [52]

    EU2-Net: A Parameter Efficient Ensemble Model With Attention-Aided Triple Feature Fusion for Tumor Segmentation in Breast Ultrasound Images,

    A. Roy et al., "EU2-Net: A Parameter Efficient Ensemble Model With Attention-Aided Triple Feature Fusion for Tumor Segmentation in Breast Ultrasound Images," IEEE Transactions on Instrumentation and Measurement, vol. 73, no. -, p. 7, 2024

    2024