Skip to content
2000
Volume 21, Issue 1
  • ISSN: 1573-4056
  • E-ISSN: 1875-6603
file format text download EPUB
file format pdf download PDF
side by side viewer icon HTML

Abstract

Background

Semantic segmentation algorithms are essential for identifying and segmenting human organs and lesions in medical images. However, as U-Net variants enhance segmentation accuracy, they often increase in parameter count, demanding more sophisticated and costly hardware for training.

Objective

This study aims to introduce a lightweight U-Net that optimizes the trade-off between network parameters and segmentation accuracy, while fully leveraging the encoder's feature extraction capabilities.

Methods

We propose a lightweight full-encoder U-shaped network, termed LFE-UNet, which employs full-encoder skip connections, encompassing all encoder layers. This model is designed with a reduced number of basic channels—specifically, 8 instead of the typical 64 or 32—to achieve a more efficient architecture.

Results

The LFE-UNet, when integrated with ResNet34, achieved a score of 0.97385 on the ISBI LiTS 2017 liver dataset. For the BraTS 2018 brain tumor dataset, it obtained 0.87510, 0.93759, 0.87301, and 0.81469 on average, WT, TC, and ET, respectively. The paper also discusses the impact of varying basic channel numbers and encoder layer counts on the network's parameter efficiency, as well as the model's robustness to different levels of Gaussian noise in images and salt and pepper noise in labels. Additionally, the influence of different loss functions is explored.

Conclusion

The LFE-UNet proves that high segmentation accuracy can be attained with a markedly lower parameters, fully utilizing the full-scale encoder's feature extraction. It also highlights the significance of loss function selection and the effects of noise on segmentation accuracy.

2025 © 2025 The Author(s). Published by Bentham Science Publisher. The Author(s)
This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International Public License (CC-BY 4.0), a copy of which is available at: https://creativecommons.org/licenses/by/4.0/legalcode. This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Loading

Article metrics loading...

/content/journals/cmir/10.2174/0115734056370555250426140155
2025-05-08
2025-09-04

  • From This Site

    /content/journals/cmir/10.2174/0115734056370555250426140155
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

/deliver/fulltext/cmir/21/1/CMIR-21-E15734056370555.html?itemId=/content/journals/cmir/10.2174/0115734056370555250426140155&mimeType=html&fmt=ahah

References

  1. RonnebergerO. FischerP. BroxT. U-Net: Convolutional networks for biomedical image segmentation.Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015Springer, Cham, 2018, vol. 9351, pp. 234-241.10.1007/978‑3‑319‑24574‑4_28
     [Google Scholar]
  2. ZhouZ. Rahman SiddiqueeM.M. TajbakhshN. LiangJ. UNet++: A nested U-Net architecture for medical image segmentation.Lect. Notes Comput. Sci.20181104531110.1007/978‑3‑030‑00889‑5_132613207
     [Google Scholar]
  3. HuangH. LinL. TongR. Unet 3+: A full-scale connected unet for medical image segmentation.IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)Barcelona, Spain, 2020 May 4-8, pp. 1055-1059.10.1109/ICASSP40776.2020.9053405
     [Google Scholar]
  4. Krithika alias AnbuDeviM. SuganthiK. Review of semantic segmentation of medical images using modified architectures of UNET.Diagnostics20221212306410.3390/diagnostics1212306436553071
     [Google Scholar]
  5. PunnN.S. AgarwalS. Modality specific U-Net variants for biomedical image segmentation: A survey.Artif. Intell. Rev.20225575845588910.1007/s10462‑022‑10152‑135250146
     [Google Scholar]
  6. KiranI. RazaB. IjazA. KhanM.A. DenseRes-Unet: Segmentation of overlapped/clustered nuclei from multi organ histopathology images.Comput. Biol. Med.202214310526710.1016/j.compbiomed.2022.10526735114445
     [Google Scholar]
  7. YinS. DengH. XuZ. ZhuQ. ChengJ. SD-UNet: A novel segmentation framework for CT images of lung infections.Electronics2022111130[http://dx.doi.org/10.3390/electronics11010130].10.3390/electronics11010130
     [Google Scholar]
  8. GuoC. SzemenyeiM. YiY. SA-UNet: Spatial attention U-Net for retinal vessel segmentation.25th International Conference on Pattern Recognition (ICPR)Milan, Italy, 10-15 January 2021, pp. 1236-1242.10.1109/ICPR48806.2021.9413346
     [Google Scholar]
  9. BougourziF. DistanteC. DornaikaF. Taleb-AhmedA. PDAtt-unet: Pyramid dual-decoder attention unet for covid-19 infection segmentation from CT-scans.Med. Image Anal.20238610279710.1016/j.media.2023.10279736966605
     [Google Scholar]
  10. MajiD SigedarP SinghM. Attention Res-UNet with guided decoder for semantic segmentation of brain tumors.Biomed Signal Process Control202271Part A10307710.1016/j.bspc.2021.103077
     [Google Scholar]
  11. CaoH. WangY. ChenJ. Swin-Unet: Unet-like pure transformer for medical image segmentation.Computer Vision – ECCV 2022 Workshops.ChamSpringer20231380320521810.1007/978‑3‑031‑25066‑8_9
     [Google Scholar]
  12. YanX. TangH. SunS. AFTer-UNet: Axial fusion transformer UNet for medical image segmentation.IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)Waikoloa, HI, USA, 03-08 January 2022, pp. 3270-3280.10.1109/WACV51458.2022.00333
     [Google Scholar]
  13. ChenD. YangW. WangL. TanS. LinJ. BuW. PCAT-UNet: UNet-like network fused convolution and transformer for retinal vessel segmentation.PLoS One2022171026268910.1371/journal.pone.026268935073371
     [Google Scholar]
  14. KhanR.A. LuoY. WuF.X. RMS-UNet: Residual multi-scale UNet for liver and lesion segmentation.Artif. Intell. Med.202212410223110.1016/j.artmed.2021.10223135115126
     [Google Scholar]
  15. ZhangY. ChenZ. YangX. Light-M: An efficient lightweight medical image segmentation framework for resource-constrained IoMT.Comput. Biol. Med.202417010808810.1016/j.compbiomed.2024.10808838320339
     [Google Scholar]
  16. HuY. MuN. LiuL. ZhangL. JiangJ. LiX. Slimmable transformer with hybrid axial-attention for medical image segmentation.Comput. Biol. Med.202417310837010.1016/j.compbiomed.2024.10837038564854
     [Google Scholar]
  17. DengY. WangH. HouY. LiangS. ZengD. LFU-Net: A lightweight UNet with full skip connections for medical image segmentation.Curr. Med. Imaging202319422062220629610.2174/157340561866622062215485335733312
     [Google Scholar]
  18. DengY HouY YanJ. ELU-Net: An efficient and lightweight U-net for medical image segmentation.IEEE Access202210359323594110.1109/ACCESS.2022.3163711
     [Google Scholar]
  19. ZhaoH. ShiJ. QiX. Pyramid scene parsing network.IEEE Conference on Computer Vision and Pattern Recognition (CVPR)Honolulu, HI, USA, 2017 July 21-26, pp. 6230-6239.10.1109/CVPR.2017.660
     [Google Scholar]
  20. ChenL.C. PapandreouG. KokkinosI. MurphyK. YuilleA.L. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs.IEEE Trans. Pattern Anal. Mach. Intell.201840483484810.1109/TPAMI.2017.269918428463186
     [Google Scholar]
  21. ChenL.C. PapandreouG. SchroffF. Rethinking atrous convolution for semantic image segmentation.IEEE Conference on Computer Vision and Pattern Recognition (CVPR)Honolulu, HI, USA, 2017 July 21-26, pp. 1-14.10.48550/arXiv.1706.05587
     [Google Scholar]
  22. ChenL.C. ZhuY. PapandreouG. Encoder-decoder with atrous separable convolution for semantic image segmentation.European conference on computer vision (ECCV)SpringerCham20181121183385110.1007/978‑3‑030‑01234‑2_49
     [Google Scholar]
  23. OktayO. SchlemperJ. FolgocL.L. Attention u-net: Learning where to look for the pancreas.IEEE/CVF Conference on Computer Vision and Pattern RecognitionSalt Lake City, UT, USA, 2018 June 18-22, pp. 1-10..10.48550/arXiv.1804.03999
     [Google Scholar]
  24. OuJ. JiangL. BaiT. ZhanP. LiuR. XiaoH. ResTransUnet: An effective network combined with Transformer and U-Net for liver segmentation in CT scans.Comput. Biol. Med.202417710862510.1016/j.compbiomed.2024.10862538823365
     [Google Scholar]
  25. McKinleyR. MeierR. WiestR. Ensembles of densely-connected CNNs with label-uncertainty for brain tumor segmentation.International MICCAI Brainlesion WorkshopSpringerCham20191138445646510.1007/978‑3‑030‑11726‑9_40
     [Google Scholar]
  26. ZhouC. DingC. WangX. LuZ. TaoD. One-pass multi-task networks with cross-task guided attention for brain tumor segmentation.IEEE Trans. Image Process.2020294516452910.1109/TIP.2020.297351032086210
     [Google Scholar]
  27. KermiA. MahmoudiI. KhadirM.T. Deep convolutional neural networks using U-Net for automatic brain tumor segmentation in multimodal MRI volumes.International MICCAI Brainlesion WorkshopSpringerCham201911384374810.1007/978‑3‑030‑11726‑9_4
     [Google Scholar]
  28. AlbiolA. AlbiolA. AlbiolF. Extending 2D deep learning architectures to 3D image segmentation problems.International MICCAI Brainlesion WorkshopGranada, Spain, 2018, pp. 73-82.10.1007/978‑3‑030‑11726‑9_7
     [Google Scholar]
  29. FengX. TustisonN. MeyerC. Brain tumor segmentation using an ensemble of 3D U-Nets and overall survival prediction using radiomic features.International MICCAI Brainlesion WorkshopSpringerCham20191138427928810.1007/978‑3‑030‑11726‑9_25
     [Google Scholar]
  30. AhmadP HaiJ RoobaeaA. MH MH UNet: A multi-scale hierarchical based architecture for medical image segmentation.IEEE Access2021914838414840810.1109/ACCESS.2021.3122543
     [Google Scholar]
  31. MyronenkoA. 3D MRI brain tumor segmentation using autoencoder regularization.International MICCAI Brainlesion WorkshopSpringerCham20191138431132010.1007/978‑3‑030‑11726‑9_28
     [Google Scholar]
  32. IsenseeF. KickingerederP. WickW. No new-net.International MICCAI Brainlesion WorkshopSpringerCham20191138423424410.1007/978‑3‑030‑11726‑9_21
     [Google Scholar]
  33. LiT. LiX. WangJ. WenJ. LuH. HsiehJ. LiangZ. Nonlinear sinogram smoothing for low-dose X-ray CT.IEEE Trans. Nucl. Sci.20045152505251310.1109/TNS.2004.834824
     [Google Scholar]
  34. KarimiD. DouH. WarfieldS.K. GholipourA. Deep learning with noisy labels: Exploring techniques and remedies in medical image analysis.Med. Image Anal.20206510175910.1016/j.media.2020.10175932623277
     [Google Scholar]
  35. DengY. GuF. ZengD. LuJ. LiuH. HouY. ZhangQ. An effective U-Net and BiSeNet complementary network for spine segmentation.Biomed. Signal Process. Control20248910568210.1016/j.bspc.2023.105682
     [Google Scholar]
/content/journals/cmir/10.2174/0115734056370555250426140155
Loading
LFE-UNet: A Lightweight Full-Encoder U-shaped Network for Efficient Semantic Segmentation in Medical Imaging
Curr. Med. Imaging 21, E15734056370555 (2025); https://doi.org/10.2174/0115734056370555250426140155
/content/journals/cmir/10.2174/0115734056370555250426140155
/content/journals/cmir/10.2174/0115734056370555250426140155
Loading

Data & Media loading...


  • Article Type:     Research Article
Keyword(s): Full-encoder skip connections; Lightweight; Medical imaging; Semantic segmentation; U-Net

Most Cited Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error
Please enter a valid_number test