Lightweight deep training network for lymph nodes segmentation from head and neck CT images

Fan Lu; Xiao‐Long Li; Binbin Jiang; Zhongyi Zhang; Caiyin Tang; Quan Li; Jing Cai; Tao Peng

doi:10.1002/mp.70123

journal article Nov 01, 2025

Lightweight deep training network for lymph nodes segmentation from head and neck CT images

Fan Lu

Xiao‐Long Li Binbin Jiang

Zhongyi Zhang

Caiyin Tang Quan Li

Jing Cai

Tao Peng

Medical Physics Vol. 52 No. 11 · Wiley

View at Publisher Save 10.1002/mp.70123

Abstract

Abstract

Background
Accurate lymph node (LN) segmentation is highly beneficial for diagnosing and treating head and neck diseases. However, because of the varying sizes and complex shapes of LNs from the head and neck, as well as their blurred boundaries with surrounding tissues in computed tomography (CT) images, it is difficult for physicians to manually identify the region of interest (ROI). Although existing 3D‐volumetric‐convolution‐based methods play an important role in LN boundary extraction, they suffer from high computational complexity.

Purpose
To tackle these issues, we develop an efficient and lightweight volumetric convolutional neural network, named LNSNet, for the LN segmentation from the head and neck region.

Methods
Our LNSNet presented a 3D Volume Block, which mainly combines Volumetric Partial Convolution (VPConv) with point‐wise convolution to decrease computational complexity and parameter count. In addition, both a Lightweight Boundary Enhancement Module (LBEM) and a depthwise separable convolution are added to the bottom of LNSNet to improve the accuracy of LN segmentation.

Results
678 3D LNs extracted from 123 patients with head and neck cancer were used for evaluation. We trained the model using 5‐fold cross‐validation and tested it on an independent test set. Our model had fewer parameters and lower computational complexity than some state‐of‐the‐art models, with a Dice Similarity Coefficient (DSC) of up to 73.81% and the Average Surface Distance (ASD) and 95th percentile‘ Hausdorff Distance (HD95) are only 0.92 and 2.52 mm, respectively.

Conclusions
LNSNet improves computational efficiency and robustness by reducing parameter count and complexity, making it more attractive in practical applications.

Topics

No keywords indexed for this article. Browse by subject →

References

43

[1]

10.1148/radiol.220329

[2]

10.1016/j.compmedimag.2023.102312

[3]

10.1016/j.radonc.2023.109870

[4]

10.1016/j.compmedimag.2025.102535

[5]

10.3390/cancers16142538

[6]

10.1259/dmfr.20210515

[7]

10.1016/j.cmpb.2022.106752

[8]

10.1016/j.bspc.2024.106519

[9]

10.1016/j.patcog.2022.108890

[10]

10.1016/j.oraloncology.2024.106796

[11]

10.1016/j.patcog.2023.109925

[12]

U-Net: Convolutional Networks for Biomedical Image Segmentation

Olaf Ronneberger, Philipp Fischer, Thomas Brox

Lecture Notes in Computer Science 10.1007/978-3-319-24574-4_28

[13]

Li H (2022)

[14]

ÇiçekÖ AbdulkadirA LienkampSS BroxT RonnebergerO.3D U‐Net: learning dense volumetric segmentation from sparse annotation. In:Proceedings of the International Conference on Medical Image Computing and Computer‐Assisted Intervention.Springer;2016:424‐432. 10.1007/978-3-319-46723-8_49

[15]

10.1117/12.2287066

[16]

10.1016/j.ebiom.2020.102780

[17]

10.3389/fnins.2024.1448294

[18]

10.1109/tmi.2022.3144274

[19]

MengZ ZhuY FanX TianJ NieF WangK.CEUSegNet: a cross‐modality lesion segmentation network for contrast‐enhanced ultrasound. In:IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE;2022:1‐5. 10.1109/isbi52829.2022.9761594

[20]

10.1109/tmi.2021.3062902

[21]

YanK JinD GuoD et al.Anatomy‐aware lymph node detection in chest CT using implicit station stratification. In:Proceedings of the International Conference on Medical Image Computing and Computer‐Assisted Intervention.Springer;2023:299‐310. 10.1007/978-3-031-47425-5_27

[22]

10.1109/jbhi.2020.3008759

[23]

10.1109/jbhi.2024.3350245

[24]

10.1186/s41747-023-00360-x

[25]

Engelson S (2024)

[26]

Cabrera DF (2021)

[27]

10.3390/cancers14225501

[28]

10.1016/j.compmedimag.2020.101851

[29]

10.1016/j.ijrobp.2021.02.004

[30]

LiuZ MaoH WuCY FeichtenhoferC DarrellT XieS.A ConvNet for the 2020s. In:IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).IEEE;2022:11976‐11986. 10.1109/cvpr52688.2022.01167

[31]

LiuZ LinY CaoY et al.Swin transformer: hierarchical vision transformer using shifted windows. In:IEEE/CVF International Conference on Computer Vision (ICCV).IEEE;2021:10012‐10022. 10.1109/iccv48922.2021.00986

[32]

ChenJ KaoSh HeH et al.Run don't walk: chasing higher FLOPS for faster neural networks. In:IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).IEEE;2023:12021‐12031. 10.1109/cvpr52729.2023.01157

[33]

MaasAL HannunAY NgAY.Rectifier nonlinearities improve neural network acoustic models. In:Proceedings of the 30 th International Conference on Machine Learning. JMLR: W&CP;2013;30:3Atlanta.

[34]

Deep Networks with Stochastic Depth

Gao Huang, Yu Sun, Zhuang Liu et al.

Lecture Notes in Computer Science 10.1007/978-3-319-46493-0_39

[35]

IslamM LiY RenH.Learning where to look while tracking instruments in robot‐assisted surgery. In:Proceedings of the International Conference on Medical Image Computing and Computer‐Assisted Intervention.Springer;2019:412‐420. 10.1007/978-3-030-32254-0_46

[36]

LoshchilovI HutterF.Decoupled weight decay regularization. In:International Conference on Learning Representations (ICLR).2019.

[37]

10.1186/s40644-022-00506-x

[38]

10.1016/j.diii.2021.04.009

[39]

10.1016/j.compmedimag.2023.102312

[40]

10.1109/tmi.2019.2913184

[41]

LeeHH BaoS HuoY LandmanBA.3D UX‐Net: a large kernel volumetric convnet modernizing hierarchical transformer for medical image segmentation. In:International Conference on Learning Representations (ICLR).2023.

[42]

V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation

Fausto Milletari, Nassir Navab, Seyed-Ahmad Ahmadi

2016 Fourth International Conference on 3D Vision... 10.1109/3dv.2016.79

[43]

HatamizadehA TangY NathV et al.UNETR: transformers for 3D medical image segmentation. In:IEEE/CVF Winter Conference on Applications of Computer Vision (WACV).IEEE;2022:574‐584. 10.1109/wacv51458.2022.00181

Metrics

0

Citations

43

References

Details

Published: Nov 01, 2025
Vol/Issue: 52(11)
License: View

Authors

F

Fan Lu

School of Future Science and Engineering Soochow University Suzhou Jiangsu China

X

Xiao‐Long Li

Department of Ultrasound, Institute of Ultrasound in Medicine and Engineering, Zhongshan Hospital Fudan University Shanghai China

B

Binbin Jiang

Key laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), department of Ultrasonography Peking University Cancer Hospital & Institute Beijing China

Z

Zhongyi Zhang

Key laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), department of Ultrasonography Peking University Cancer Hospital & Institute Beijing China

C

Caiyin Tang

Department of Radiology Taizhou People's Hospital Affiliated to Nanjing Medical University Taizhou China

Q

Quan Li

Center of Stomatology the Second Affiliated Hospital of Soochow University Suzhou China

J

Jing Cai

Department of Health Technology and Informatics The Hong Kong Polytechnic University Hong Kong China

T

Tao Peng

School of Future Science and Engineering Soochow University Suzhou Jiangsu China; Department of Health Technology and Informatics The Hong Kong Polytechnic University Hong Kong China

Cite This Article

Fan Lu, Xiao‐Long Li, Binbin Jiang, et al. (2025). Lightweight deep training network for lymph nodes segmentation from head and neck CT images. Medical Physics, 52(11). https://doi.org/10.1002/mp.70123

Lightweight deep training network for lymph nodes segmentation from head and neck CT images

You May Also Like