[1]王佳浩,王宝珠,郭志涛,等.多感受野的轻量化YOLOv4用于检测肺结核[J].中国医学物理学杂志,2022,39(9):1119-1127.[doi:DOI:10.3969/j.issn.1005-202X.2022.09.011]
　WANG Jiahao,WANG Baozhu,GUO Zhitao,et al.Lightweight YOLOv4 with multi-receptive fields for detection of pulmonary tuberculosis[J].Chinese Journal of Medical Physics,2022,39(9):1119-1127.[doi:DOI:10.3969/j.issn.1005-202X.2022.09.011]
点击复制

多感受野的轻量化YOLOv4用于检测肺结核()
分享到：

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

卷:
39卷
期数:
2022年第9期
页码:
1119-1127
栏目:
医学影像物理
出版日期:
2022-11-02

文章信息/Info

Title:
Lightweight YOLOv4 with multi-receptive fields for detection of pulmonary tuberculosis
文章编号:
1005-202X（2022）09-1119-09
作者:
王佳浩; 王宝珠; 郭志涛; 王京华
河北工业大学电子信息工程学院， 天津 300401
Author(s):
WANG Jiahao;  WANG Baozhu;  GUO Zhitao;  WANG Jinghua
School of Electronic and Information Engineering, Hebei University of Technology, Tianjin 300401, China
关键词:
肺结核; YOLOv4; MobileNetv3; 多感受野
Keywords:
Keywords: pulmonary tuberculosis YOLOv4 MobileNetv3 multi-receptive field
分类号:
R318；R521
DOI:
DOI:10.3969/j.issn.1005-202X.2022.09.011
文献标志码:
A
摘要:
肺结核疾病特征错综复杂，人工筛查成本较高，缺少规范的数据集。当前基于卷积神经网络的检测模型结构复杂、参数量大且检测精度有待进一步提高，为此提出一种改进的轻量化YOLOv4的肺结核检测模型。首先选取300例实际病例，制作一套规范的数据集，用于评估模型的性能；随后通过残差通道注意力模块改进MobileNetv3的结构，并作为YOLOv4的主干提取器，进一步减少参数量并融合上下文信息；然后在主干提取器的3个有效特征层后加入多感受野模块，有效增强低特征层的信息提取能力并降低对小型肺结核病灶的漏检率；最后，将以上改进的模块与YOLOv4的多尺度结构相结合，构建一种多感受野的轻量化YOLOv4的肺结核检测模型。与原始YOLOv4相比，该模型的参数量减少了约47%，平均精准度（mAP）值提升至96.60%，漏检率降低至6%，验证该模型能有效辅助影像科医师诊断肺结核。
Abstract:
Abstract: The characteristics of pulmonary tuberculosis are complex, with the high cost of manual screening, lack of standardized data sets. The current detection model based on convolution neural network has intricate structure, large number of parameters and detection accuracy needs to be further ameliorated. Therefore, an improved lightweight YOLOv4 model is proposed for pulmonary tuberculosis detection. A standardized dataset is constructed using 300 actual cases for evaluating the performance of the model. MobileNetv3 improved with residual channel attention module is used as the backbone extractor of YOLOv4 for further decreasing the number of parameters and fusing context information. Then the multi-receptive field module is added after the 3 effective feature layers of the backbone extractor, which effectively enhances the information extraction ability of the low feature layer and reduces the missed etection rate of small pulmonary tuberculosis lesions. The above improved modules were combined with the multi-scale structure of YOLOv4 to construct a lightweight YOLOv4 model with multi-receptive field for pulmonary tuberculosis detection.Compared with the original YOLOv4, the proposed model reduces the number of parameters of the model by about 47%, elevates the mAP value to 96.60%, and decreases the missed detection rate to 6%. It is verified that lightweight YOLOv4 with multi-receptive fields can effectively assist radiologists in the diagnosis of pulmonary tuberculosis.

相似文献/References:

[1]安冬会,杨俊潇,屈亚林,等. 能谱CT在肺部磨玻璃结节良恶性鉴别中的应用价值[J].中国医学物理学杂志,2018,35(10):1160.[doi:DOI:10.3969/j.issn.1005-202X.2018.10.009]
　AN Donghui,YANG Junxiao,QU Yalin,et al. Application value of energy spectrum CT in the differential diagnosis of benign and malignant pulmonary ground-glass nodule[J].Chinese Journal of Medical Physics,2018,35(9):1160.[doi:DOI:10.3969/j.issn.1005-202X.2018.10.009]
[2]杨朝生,陈慧,孙雪皎,等.结核感染T细胞斑点试验与肺部CT表现的相关性[J].中国医学物理学杂志,2020,37(9):1155.[doi:10.3969/j.issn.1005-202X.2020.09.014]
　YANG Chaosheng,CHEN Hui,SUN Xuejiao,et al.Correlation between T-SPOT.TB assay and pulmonary CT findings in pulmonary tuberculosis[J].Chinese Journal of Medical Physics,2020,37(9):1155.[doi:10.3969/j.issn.1005-202X.2020.09.014]
[3]翟安,邓平,贝承丽,等.胸部物理治疗联合纤维支气管镜灌洗对肺结核的效果[J].中国医学物理学杂志,2021,38(8):1023.[doi:DOI:10.3969/j.issn.1005-202X.2021.08.020]
　ZHAI An,DENG Ping,BEI Chengli,et al.Therapeutic effect of chest physiotherapy combined with fiberoptic bronchoscopy lavage on pulmonary tuberculosis[J].Chinese Journal of Medical Physics,2021,38(9):1023.[doi:DOI:10.3969/j.issn.1005-202X.2021.08.020]
[4]郑秋婷,郭琳,夏丽,等.基于CT的空洞特征在耐药肺结核的影像诊断价值[J].中国医学物理学杂志,2024,41(4):413.[doi:DOI:10.3969/j.issn.1005-202X.2024.04.003]
　ZHENG Qiuting,GUO Lin,XIA Li,et al.Diagnostic value of CT cavity features in drug-resistant pulmonary tuberculosis[J].Chinese Journal of Medical Physics,2024,41(9):413.[doi:DOI:10.3969/j.issn.1005-202X.2024.04.003]
[5]谢浩杰,鲁明丽,张陈,等.基于深度学习的肺结核检测综述[J].中国医学物理学杂志,2024,41(7):918.[doi:DOI:10.3969/j.issn.1005-202X.2024.07.020]
　XIE Haojie,LU Mingli,ZHANG Chen,et al.Review on tuberculosis detection using deep learning[J].Chinese Journal of Medical Physics,2024,41(9):918.[doi:DOI:10.3969/j.issn.1005-202X.2024.07.020]
[6]王京华,袁金丽,郭志涛,等.改进的YOLOv4算法在肺结核检测中的应用研究[J].中国医学物理学杂志,2023,40(1):113.[doi:DOI:10.3969/j.issn.1005-202X.2023.01.019]
　WANG Jinghua,YUAN Jinli,GUO Zhitao,et al.Application of improved YOLOv4 algorithm in the detection of pulmonary tuberculosis[J].Chinese Journal of Medical Physics,2023,40(9):113.[doi:DOI:10.3969/j.issn.1005-202X.2023.01.019]

备注/Memo

备注/Memo:
【收稿日期】2021-11-23 【基金项目】国家自然科学基金（61801164） 【作者简介】王佳浩，硕士，研究方向：智能信息处理、计算机视觉、机器学习、医疗图像处理，E-mail: wangjiahaolad777@163.com
更新日期/Last Update: 2022-09-27