[1]陈星月,贾子彦,李青,等.基于改进YOLOv5s的免疫组化阳性细胞计数方法[J].中国医学物理学杂志,2025,42(2):167-174.[doi:DOI:10.3969/j.issn.1005-202X.2025.02.005]
　CHEN Xingyue,JIA Ziyan,LI Qing,et al.Improved YOLOv5s based method for immunohistochemically positive cell counting[J].Chinese Journal of Medical Physics,2025,42(2):167-174.[doi:DOI:10.3969/j.issn.1005-202X.2025.02.005]
点击复制

基于改进YOLOv5s的免疫组化阳性细胞计数方法()
分享到：

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
42
期数:
2025年第2期
页码:
167-174
栏目:
医学影像物理
出版日期:
2025-01-20

文章信息/Info

Title:
Improved YOLOv5s based method for immunohistochemically positive cell counting
文章编号:
1005-202X（2025）02-0167-08
作者:
陈星月1; 贾子彦1; 李青2; 张大川2; 潘玲佼1; 沈大伟1
1.江苏理工学院电气信息工程学院， 江苏 常州 213001； 2.常州市第一人民医院病理科， 江苏 常州 213004
Author(s):
CHEN Xingyue1;  JIA Ziyan1;  LI Qing2;  ZHANG Dachuan2;  PAN Lingjiao1;  SHEN Dawei1
1. School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China 2. Department of Pathology, Changzhou First Peoples Hospital, Changzhou 213004, China
关键词:
阳性细胞; 目标检测; YOLOv5s; 免疫组织化学; 生存预测
Keywords:
Keywords: positive cell target detection YOLOv5s immunohistochemistry survival prediction
分类号:
R318；TP183
DOI:
DOI:10.3969/j.issn.1005-202X.2025.02.005
文献标志码:
A
摘要:
目的：提出一种基于改进YOLOv5s的免疫组化阳性细胞计数方法。方法：首先，针对阳性细胞的小目标特征，增加小目标检测层细化特征提取；其次，在颈部网络中使用双向加权特征金字塔网络替换PANet结构，实现多尺度特征融合；再次，增加坐标注意力机制CA模块，使模型更加关注小目标特征；最后，用EIoU损失函数替换原有的GIoU，增强模型检测性能。结果：在自建的免疫组化图像数据集上进行训练，改进后的模型平均精确率为89.3%，较原模型提升4.0%，且优于其他主流目标检测模型。同时，基于该方法构建的5年生存预测模型平均准确率为76.8%，平均AUC为0.81，表明模型具有较好的预测能力。结论：本研究模型能够快速检测免疫组化阳性细胞数量并有效地辅助医生进行生存预测工作。
Abstract:
Abstract: Objective To propose a novel method for immunohistochemically positive cell counting based on the improved YOLOv5s. Methods Regarding the small target characteristics of positive cells, a small target detection layer was added to refine feature extraction. Then, a bidirectional weighted feature pyramid network was used to replace path aggregation network (PANet) in the neck network for realizing multi-scale feature fusion. Additionally, the method used coordinate attention mechanism to make the model pay more attention to small target characteristics, and replaced the original GIoU with EIoU loss function for enhancing the detection performance. Results The model was trained on the self-built immunohistochemical image dataset. The average accuracy of the improved model was 89.3%, which was 4.0% higher than the original model and surpassed mainstream target detection models. The 5-year survival prediction model constructed with the method achieved an average accuracy of 76.8% and an average area under the curve of 0.81, demonstrating its superior prediction ability. Conclusion The proposed model can quickly detect the number of immunohistochemically positive cells and effectively assist doctors in survival prediction.

相似文献/References:

[1]梁智欣,邓戈龙,魏夏平,等.长尾分布下的PET-CT肺癌图像的深度学习网络检测与定位探索[J].中国医学物理学杂志,2022,39(12):1473.[doi:DOI:10.3969/j.issn.1005-202X.2022.12.004]
　LIANG Zhixin,DENG Gelong,WEI Xiaping,et al.Deep learning network for lung cancer detection and localization in PET-CT images under long-tailed distribution[J].Chinese Journal of Medical Physics,2022,39(2):1473.[doi:DOI:10.3969/j.issn.1005-202X.2022.12.004]
[2]米吾尔依提·海拉提,热娜古丽·艾合麦提尼亚孜,卡迪力亚·库尔班,等.基于改进YOLOv7的肝囊型包虫病超声图像小病灶检测[J].中国医学物理学杂志,2024,41(3):299.[doi:DOI:10.3969/j.issn.1005-202X.2024.03.006]
　HAILATI Miwueryiti,AIHEMAITINIYAZI Renaguli,KUERBAN Kadiliya,et al.Small lesion detection in ultrasound images of hepatic cystic echinococcosis based on improved YOLOv7[J].Chinese Journal of Medical Physics,2024,41(2):299.[doi:DOI:10.3969/j.issn.1005-202X.2024.03.006]
[3]贺斌,高永彬.基于非局部卷积和卷积注意力模块的眩晕眼震诊断方法[J].中国医学物理学杂志,2024,41(5):571.[doi:DOI:10.3969/j.issn.1005-202X.2024.05.007]
　HE Bin,GAO Yongbin.Diagnostic methods for nystagmus in vertigo based on non-local convolution and convolutional block attention module[J].Chinese Journal of Medical Physics,2024,41(2):571.[doi:DOI:10.3969/j.issn.1005-202X.2024.05.007]
[4]刘静妮,盛玉武,赵长秀,等.基于卷积神经网络的X线片下肢关节角度识别算法[J].中国医学物理学杂志,2024,41(8):996.[doi:DOI:10.3969/j.issn.1005-202X.2024.08.012]
　LIU Jingni,SHENG Yuwu,ZHAO Changxiu,et al.Lower limb joint angle calculation algorithm based on convolutional neural network in X-ray films[J].Chinese Journal of Medical Physics,2024,41(2):996.[doi:DOI:10.3969/j.issn.1005-202X.2024.08.012]
[5]张琼,杭益柳,邱建林,等.高效注意力金字塔网络在肺结节检测的应用[J].中国医学物理学杂志,2024,41(11):1361.[doi:DOI:10.3969/j.issn.1005-202X.2024.11.007]
　ZHANG Qiong,HANG Yiliu,et al.Efficient attention feature pyramid network for pulmonary nodule detection[J].Chinese Journal of Medical Physics,2024,41(2):1361.[doi:DOI:10.3969/j.issn.1005-202X.2024.11.007]

备注/Memo

备注/Memo:
【收稿日期】2024-07-13 【基金项目】国家自然科学基金（62001196）；江苏省第六期333高层次人才培养工程（2022-2）；常州市5G+工业互联网融合应用重点实验室（CM20223015）；常州市应用基础研究（CJ20220064，CJ20220059） 【作者简介】陈星月，硕士研究生，研究方向：医学图像处理、目标检测，E-mail: 643395909@qq.com 【通信作者】贾子彦，副教授，研究方向：可见光通信、机器视觉、5G研究，E-mail: jiaziyan@jsut.edu.cn
更新日期/Last Update: 2025-01-22