首页 > 论文 > 光学学报 > 40卷 > 18期(pp:1810003--1)

基于部分注释CT图像的自监督迁移学习肺结节分类

Self-Supervised Transfer Learning of Pulmonary Nodule Classification Based on Partially Annotated CT Images

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

深度学习模型训练时需要大量的注释样本,但在医学领域注释数据难以获取。针对此问题,提出了一种结合部分注释数据的自监督学习算法,以提高3D肺结节的分类性能。在传统自监督训练的网络结构基础上,设计了一种多任务学习的网络结构,以同时利用医学图像处理任务中大量未注释数据和少量注释数据。通过先训练未注释数据然后加入注释数据继续训练的方式,实现了注释数据与未注释数据间部分网络结构和参数的共享。相较于传统自监督学习方法,所提算法在保证模型泛化能力的同时能够学习到更多与肺结节相关的鉴别特征,因此将模型迁移学习用于肺结节分类时也能表现出更佳的性能。所提算法在公开数据集LIDC-IDRI上的分类准确率达0.886,曲线下面积(AUC)值达0.929,实验结果表明,所提算法能够有效提升肺结节的分类性能。

Abstract

The process of training a deep learning model requires many annotation samples, even though annotation data is difficult to obtain in the medical field. A self-supervised learning algorithm combined with partial annotation data is proposed as a solution to this problem, in order to improve classification performance of 3D pulmonary nodules. Based on the traditional self-supervised training network structure, a multitask learning network structure is designed to address a large amount of unannotated data and a small amount of annotated data obtained from medical image processing tasks. First, the proposed algorithm trains the unannotated data, and then explores the annotation data to continuously train the model. Thus, this algorithm manages to share partial network structures and parameters between the annotated and unannotated data. Compared to traditional self-supervised learning methods, the proposed algorithm can learn to recognize the discriminant features of pulmonary nodules to ensure the model''s capacity to generalize, therefore, model transfer learning can also perform better when applied to the classification of pulmonary nodules. The classification accuracy of the proposed algorithm on LIDC-IDRI dataset is 0.886, and the area under the curve (AUC) is 0.929. The results of the investigation indicate that the proposed algorithm can effectively improve classification performance of pulmonary nodules.

广告组1 - 空间光调制器+DMD
补充资料

中图分类号:TP391

DOI:10.3788/AOS202040.1810003

所属栏目:图像处理

基金项目:中央高校基本科研业务费“医工融合项目”、重庆市科卫联合项目医学科研项目、重庆市科研院所绩效激励引导专项、中华国际医学交流基金会2019 SKY影像科研基金;

收稿日期:2020-04-17

修改稿日期:2020-06-11

网络出版日期:2020-09-01

作者单位    点击查看

黄鸿:重庆大学光电技术及系统教育部重点实验室, 重庆 400044
彭超:重庆大学光电技术及系统教育部重点实验室, 重庆 400044
吴若愚:重庆大学光电技术及系统教育部重点实验室, 重庆 400044
陶俊利:重庆大学附属肿瘤医院影像科, 重庆 400030
张久权:重庆大学附属肿瘤医院影像科, 重庆 400030

联系人作者:黄鸿(hhuang@cqu.edu.cn)

备注:中央高校基本科研业务费“医工融合项目”、重庆市科卫联合项目医学科研项目、重庆市科研院所绩效激励引导专项、中华国际医学交流基金会2019 SKY影像科研基金;

【1】Chen W Q, Zheng R S, Baade P D, et al. Cancer statistics in China, 2015 [J]. CA: a Cancer Journal for Clinicians. 1900, 66(2): 115-132.

【2】Jacobs C, van Rikxoort E M, Scholten E T, et al. Solid, part-solid, or non-solid: classification of pulmonary nodules in low-dose chest computed tomography by a computer-aided diagnosis system [J]. Investigative Radiology. 2015, 50(3): 168-173.

【3】Feng Y, Yi B S, Wu C Y, et al. Pulmonary nodule recognition based on three-dimensional convolution neural network [J]. Acta Optica Sinica. 2019, 39(6): 0615006.
冯雨, 易本顺, 吴晨玥, 等. 基于三维卷积神经网络的肺结节识别研究 [J]. 光学学报. 2019, 39(6): 0615006.

【4】Zhao Q Y, Kong P, Min J Z, et al. A review of deep learning methods for the detection and classification of pulmonary nodules [J]. Journal of Biomedical Engineering. 2019, 36(6): 1060-1068.
赵清一, 孔平, 闵建中, 等. 肺结节检测与分类的深度学习方法综述 [J]. 生物医学工程学杂志. 2019, 36(6): 1060-1068.

【5】Miao G, Li C F. Detection of pulmonary nodules CT images combined with two-dimensional and three-dimensional convolution neural networks [J]. Laser & Optoelectronics Progress. 2018, 55(5): 051006.
苗光, 李朝锋. 二维和三维卷积神经网络相结合的CT图像肺结节检测方法 [J]. 激光与光电子学进展. 2018, 55(5): 051006.

【6】Setio A A A, Jacobs C, Gelderblom J, et al. Automatic detection of large pulmonary solid nodules in thoracic CT images [J]. Medical Physics. 2015, 42(10): 5642-5653.

【7】Chen C H, Chang C K, Tu C Y, et al. Radiomic features analysis in computed tomography images of lung nodule classification [J]. PLoS One. 2018, 13(2): e0192002.

【8】Dhara A K, Mukhopadhyay S, Dutta A, et al. A combination of shape and texture features for classification of pulmonary nodules in lung CT images [J]. Journal of Digital Imaging. 2016, 29(4): 466-475.

【9】Anwar S M, Majid M, Qayyum A, et al. Medical image analysis using convolutional neural networks: a review [J]. Journal of Medical Systems. 2018, 42(11): 1-13.

【10】Nibali A, He Z, Wollersheim D. Pulmonary nodule classification with deep residual networks [J]. International Journal of Computer Assisted Radiology and Surgery. 2017, 12(10): 1799-1808.

【11】Polat H, Mehr H D. Classification of pulmonary CT images by using hybrid 3D-deep convolutional neural network architecture [J]. Applied Sciences. 2019, 9(5): 940.

【12】Tajbakhsh N, Shin J Y, Gurudu S R, et al. Convolutional neural networks for medical image analysis: fine tuning or full training? IEEE Transactions on Medical Imaging[J]. 0, 35(5): 1299-1312.

【13】Huang S, Li F F, Chen Q. Computed tomography image classification algorithm based on improved deep residual network [J]. Acta Optica Sinica. 2020, 40(3): 0310002.
黄盛, 李菲菲, 陈虬. 基于改进深度残差网络的计算断层扫描图像分类算法 [J]. 光学学报. 2020, 40(3): 0310002.

【14】Chen L, Bentley P, Mori K, et al. Self-supervised learning for medical image analysis using image context restoration [J]. Medical Image Analysis. 2019, 58: 101539.

【15】Zhou Z W, Sodha V. Rahman Siddiquee M M, et al. Models genesis: generic autodidactic models for 3D medical image analysis [M]. ∥Sheng D G, Liu T M, Peters T M, et al. Medical image computing and computer assisted intervention-MICCAI 2019. Cham: Springer. 2019, 11767: 384-393.

引用该论文

Huang Hong,Peng Chao,Wu Ruoyu,Tao Junli,Zhang Jiuquan. Self-Supervised Transfer Learning of Pulmonary Nodule Classification Based on Partially Annotated CT Images[J]. Acta Optica Sinica, 2020, 40(18): 1810003

黄鸿,彭超,吴若愚,陶俊利,张久权. 基于部分注释CT图像的自监督迁移学习肺结节分类[J]. 光学学报, 2020, 40(18): 1810003

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF