Research on Segmentation Method of Medical Nuclear Image based on U-Net++

GONG Zikun

doi:10.11907/rjdk.251053

您当前的位置：

首页 >

文章列表页 >

Research on Segmentation Method of Medical Nuclear Image based on U-Net++

Graphic and Image Processing | 更新时间：2026-04-30

- Research on Segmentation Method of Medical Nuclear Image based on U-Net++
- Software Guide Vol. 25, Issue 4, Pages: 155-161(2026)
- 作者机构：
  
  南京邮电大学通达学院，江苏扬州 225127
- 作者简介：
- 基金信息：
- DOI：10.11907/rjdk.251053
  CLC： TP391.9
- Received：04 March 2025，
  
  Published：15 April 2026
- 稿件说明：
移动端阅览
龚梓坤.基于U-Net++的医学细胞核图像分割方法研究[J].软件导刊,2026,25(04):155-161.

GONG Zikun.Research on Segmentation Method of Medical Nuclear Image based on U-Net++[J].Software Guide,2026,25(04):155-161.
龚梓坤.基于U-Net++的医学细胞核图像分割方法研究[J].软件导刊,2026,25(04):155-161. DOI： 10.11907/rjdk.251053.

GONG Zikun.Research on Segmentation Method of Medical Nuclear Image based on U-Net++[J].Software Guide,2026,25(04):155-161. DOI： 10.11907/rjdk.251053.

摘要

针对传统方法提取医学细胞核轮廓不清晰、识别目标准确率不高的问题，提出一种基于U-Net++的医学细胞核图像分割方法，并引入降噪层对分割图像进行降噪处理

。首先，对医学细胞核图像进行合并掩码、调整图像大小的处理，形成数据集；然后，将数据集输入U-Net++进行训练，并在模型收敛后，通过剪枝操作去除冗余的参数，得到医学细胞核图像分割模型；最后，将该模型应用到医学细胞核，得到总体精度更高、分割效果更稳定的细胞核分割结果。在验证集训练中，该方法的总体精度达到84%，相比U-Net提升了19.1%，较Feedback U-Net提高了6.7%。实验结果表明，U-Net++在细胞核图像分割任务中表现出色，不仅在分割精度上明显优于其他方法，还展现出良好的泛化能力和强大的特征提取能力。

Abstract

Aiming at the problems that the traditional methods of extracting medical nuclear outline are not clear and the accuracy of target recognition is not high， a U-net ++ based medical nuclear image segmentation method is proposed and noise reduction layer is introduced to reduce the segmentation image. Firstly， the medical nuclear images are processed by merging mask and adjusting image size to form a data set. Then， the data set was input into U-NET ++ for training until convergence， and redundant parameters were removed by pruning operation to obtain a medical nuclear image segmentation model. Finally， the model is applied to medical nuclei， and the results of nuclear segmentation with higher overall precision and more stable segmentation effect are obtained. After verification set training， the overall accuracy of the proposed method can reach 84%， about 19.1% higher than U-Net， and about 6.7% higher than Feedback U-Net. The experimental results show that U-NET ++ has higher segmentation accuracy and better generalization ability than other methods in nuclear image processing， showing a strong ability to extract nuclear image features.

关键词

Keywords

references

1 ］ SIMONYAN K ， ZISSERMAN A . Very deep convolutional networks for large-scale image recognition ［C］// San Diego： International Conference on Learning Representations ， 2015 .

HE K ， ZHANG X ， REN S ， et al . Deep residual learning for image recognition ［C］// Las Vegas： Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition ， 2016 . doi: 10.1109/cvpr.2016.90 http://dx.doi.org/10.1109/cvpr.2016.90

LONG J ， SHELHAMER E ， DARRELL T . Fully convolutional networks for semantic segmentation ［C］// Boston： Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition ， 2015 . doi: 10.1109/cvpr.2015.7298965 http://dx.doi.org/10.1109/cvpr.2015.7298965

RONNEBERGER O ， FISCHER P ， BROX T . U-Net： convolutional networks for biomedical image segmentation ［C］// Munich： International Conference on Medical Image Computing and Computer-Assisted Intervention ， 2015 . doi: 10.1007/978-3-319-24574-4_28 http://dx.doi.org/10.1007/978-3-319-24574-4_28

ZHOU Z ， SIDDIQUEE M M R ， TAJBAKHSH N ， et al . UNet++： a nested u-net architecture for medical image segmentation ［C］// Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support ， 2019 ： 3 - 11 . doi: 10.1007/978-3-030-00889-5_1 http://dx.doi.org/10.1007/978-3-030-00889-5_1

HE H ， CHEN S . Automatic PET tumor segmentation based on U-Net model of pre-trained encoder ［J］. Journal of Image and Graphics ， 2020 ， 25 （ 1 ）： 171 - 179 . doi: 10.11834/jig.190058 http://dx.doi.org/10.11834/jig.190058

何慧，陈胜 . 改进预训练编码器U-Net模型的PET肿瘤自动分割［J］. 中国图象图形学报， 2020 ， 25 （ 1 ）： 171 - 179 . doi: 10.11834/jig.190058 http://dx.doi.org/10.11834/jig.190058

ZHAO T ， ZHAO Y ， WANG S ， et al . UNet++-based multi-channel speech dereverberation and distant speech recognition ［C］// Hong Kong： 12th International Symposium on Chinese Spoken Language Processing ， 2021 . doi: 10.1109/iscslp49672.2021.9362064 http://dx.doi.org/10.1109/iscslp49672.2021.9362064

LECUN Y ， BOTTOU L ， BENGIO Y ， et al . Gradient-based learning applied to document recognition ［J］. Proceedings of the IEEE ， 1998 ， 86 （ 11 ）： 2278 - 2324 . doi: 10.1109/5.726791 http://dx.doi.org/10.1109/5.726791

NAIR V ， HINTON G E . Rectified linear units improve restricted boltzmann machines ［C］// Proceedings of the 27th International Conference on Machine Learning ， 2010 ： 807 - 814 . doi: doi:http://dx.doi.org/ http://dx.doi.org/doi:http://dx.doi.org/

ZHAO X ， WANG L ， ZHANG Y ， et al . A review of convolutional neural networks in computer vision ［J］. Artificial Intelligence Review ， 2024 ， 57 （ 4 ）： 99 . doi: 10.1007/s10462-024-10721-6 http://dx.doi.org/10.1007/s10462-024-10721-6

ZHANG H M ， WU F J ， ZHENG C ， et al . Two properties of nonlinear activation function relu on mnist dataset ［J］. Journal of Hubei Normal University （Natural Science Edition）， 2024 ， 44 （ 3 ）： 1 - 7 .

张华民，吴方婧，郑超，等 . mnist数据集上非线性激活函数relu的两个性质［J］. 湖北师范大学学报（自然科学版）， 2024 ， 44 （ 3 ）： 1 - 7 .

WALTER B . Analysis of convolutional neural network image classifiers in a hierarchical max-pooling model with additional local pooling ［J］. Journal of Statistical Planning and Inference ， 2023 ， 224 ： 109 - 126 . doi: 10.1016/j.jspi.2022.11.001 http://dx.doi.org/10.1016/j.jspi.2022.11.001

HECKBERT P S . Fundamentals of texture mapping and image warping ［R］. Technical Report， No .UCB/CSD-89-516， 1989 .

WANG S ， YANG K J . Research and implementation of image scaling algorithm based on bilinear interpolation ［J］. Automation technology and applications ， 2008 ， 27 （ 7 ）： 44 - 45 .

王森，杨克俭 . 基于双线性插值的图像缩放算法的研究与实现［J］. 自动化技术与应用， 2008 ， 27 （ 7 ）： 44 - 45 .

JIAO R H ， CHU J J ， LIU J J ， et al . SC-Net： contextual information hopping linkage network for overlapping chromosome segmentation ［J］. Journal of Image and Graphics ， 2024 ， 29 （ 9 ）： 2806 - 2824 .

焦润海，褚佳杰，刘嘉骥，等 . SC-Net：用于重叠染色体分割的上下文信息跳跃连接网络［J］. 中国图象图形学报， 2024 ， 29 （ 9 ）： 2806 - 2824 .

WANG H Y . Semantic segmentation and change detection of urban remote sensing image based on RU-Net and BiUnet-Dense models ［D］. Chengdu ： Chengdu University of Technology ， 2023 .

王海英 . 基于RU-Net和BiUnet-Dense模型的城市遥感影像语义分割及变化检测［D］. 成都：成都理工大学， 2023 .

ARTIME O ， GRASSIA M ， DE D M ， et al . Robustness and resilience of complex networks ［J］. Nature Reviews Physics ， 2024 ， 6 （ 2 ）： 114 - 131 . doi: 10.1038/s42254-023-00676-y http://dx.doi.org/10.1038/s42254-023-00676-y

HSU C ， CHANG C ， CHEN T W ， et al . Brain tumor segmentation （BraTS） challenge short paper： Improving three-dimensional brain tumor segmentation using SegResNet and hybrid boundary-dice loss ［C］// International MICCAI Brainlesion Workshop ， 2021 ： 334 - 344 . doi: 10.1007/978-3-031-09002-8_30 http://dx.doi.org/10.1007/978-3-031-09002-8_30

SHIQIN G X ， YANG C R ， LIU X D ， et al . Based on attention mechanism of qin video denoising algorithms ［J/OL］. Journal of Northwest University （Natural Science Edition）， 2025-01-20 . https：//doi.org/10.16152/j.cnki.xdxbzr.2025-01-014 https://doi.org/10.16152/j.cnki.xdxbzr.2025-01-014 .

师秦高雪，杨超然，刘鑫达，等 . 基于注意力机制的秦腔视频去噪算法研究［J/OL］. 西北大学学报（自然科学版）， 2025-01-20 . https：//doi.org/10.16152/j.cnki.xdxbzr.2025-01-014 https://doi.org/10.16152/j.cnki.xdxbzr.2025-01-014 .

PUNN N S ， AGARWAL S . Modality specific U-Net variants for biomedical image segmentation： a survey ［J］. Artificial Intelligence Review ， 2022 ， 55 ： 5845 - 5889 . doi: 10.1007/s10462-022-10152-1 http://dx.doi.org/10.1007/s10462-022-10152-1

SHIBUYA E ， HOTTA K . Cell image segmentation by using feedback and convolutional LSTM ［J］. Visual Computer ， 2022 ， 38 ： 3791 - 3801 . doi: 10.1007/s00371-021-02221-3 http://dx.doi.org/10.1007/s00371-021-02221-3

ZHENG Z ， WANG P ， LIU W ， et al . Distance-IoU loss：faster and better learning for bounding box regression ［C］// Proceedings of the AAAI Conference on Artificial Intelligence ， 2020 ： 12993 - 13000 . doi: 10.1609/aaai.v34i07.6999 http://dx.doi.org/10.1609/aaai.v34i07.6999

EISUKE S ， KAZUHIRO H . Feedback U-net for cell image segmentation ［DB/OL］ . https：//arxiv.org/abs/2004.14581 https://arxiv.org/abs/2004.14581 . doi: 10.1109/cvprw50498.2020.00495 http://dx.doi.org/10.1109/cvprw50498.2020.00495

QIN C ， WU Y ， LIAO W ， et al . Improved U-Net3+ with stage residual for brain tumor segmentation ［J］. BMC Medical Imaging ， 2022 ， 22 （ 1 ）： 14 . doi: 10.1186/s12880-022-00738-0 http://dx.doi.org/10.1186/s12880-022-00738-0

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Improved Surface Defect Detection Model for Aluminum Profile Based on CSE-YOLO

A Review of Deep Learning-Based Adrenal CT Image Segmentation Research

Wide-Deep Learning Malware Detection Based on Feature Fusion

Network Intrusion Detection Model Based on Parallel Branch Feature Fusion

Cell Image Segmentation Based on Improved U-Net Model

Related Author

LIU Zilong

PANG Ming

DOU Jianming

FANG Ziyang

WANG Jinliang

XIAO Peiyao

LI Yikai

ZHANG Zhi

Related Institution

School of Mechanical and Electrical Engineering， Lanzhou Jiaotong University

School of Health Science and Engineering，University of Shanghai for Science and Technology

Department of Cardiology，Shanghai General Hospital，Shanghai Jiao Tong University School of Medicine

School of Computer Science and Technology，Taiyuan Normal University

Shanxi Key Laboratory of Intelligent Optimization Computing and Blockchain Technology

⁰