糖尿病视网膜病变分期双分支混合注意力决策网络

doi:10.13229/j.cnki.jdxbgxb20200813

摘要/Abstract

摘要：

为了缓解大规模糖尿病视网膜病变（DR）筛查需求下医疗资源不足的问题，本文提出了糖尿病视网膜病变分期双分支混合注意力决策网络（BiRAD-Net）。该网络分为特征提取和分类两个阶段：在特征提取阶段，引入混合注意力机制抑制噪声，并设计了特征分级决策网络进一步优化特征质量；在特征分类阶段，设计了双分支分类器以及对应的损失，以减缓标签数据不足带来的影响，增强分类准确性。此外，在模型训练过程中应用迁移学习技术来提高模型的精度、降低训练所需数据量。在KAGGLE数据集上的实验结果表明：本文方法对糖尿病视网膜病变的各个阶段均具有较好的诊断能力，优于其他对比方法。

关键词: 计算机应用, 深度学习, 糖尿病视网膜病变分期, 混合注意力机制, 特征分级决策网络

Abstract:

In order to address the challenges caused by limited medical resources in large-scale Diabetic Retinopathy （DR） screening process， a dual-branch hybrid attention decision net （BiRAD-Net） for DR classification is proposed. The proposed network consists of feature extraction and classification two stages. In the feature extraction stage， a hybrid attention is introduced to suppress the noise， and feature grade decision network is designed to improve feature quality. In the feature classification stage， dual-branch classifiers and corresponding dual-branch loss are designed to alleviate the impact of insufficient data with five-category labels and enhance the classification accuracy. Furthermore， the transfer learning technology is applied in the training process， together with the above steps， to improve the accuracy of model and reduce the amount of training data. Experimental results on KAGGLE dataset indicate that BiRAD-Net shows excellent diagnostic ability for all the stages of DR and preforms better than other existing comparison methods.

Key words: computer application, deep learning, diabetic retinopathy（DR） classification, hybrid attention mechanism, feature grade decision net

中图分类号:

TP391.41

欧阳继红,郭泽琪,刘思光. 糖尿病视网膜病变分期双分支混合注意力决策网络[J]. 吉林大学学报(工学版), 2022, 52(3): 648-656.

Ji-hong OUYANG,Ze-qi GUO,Si-guang LIU. Dual⁃branch hybrid attention decision net for diabetic retinopathy classification[J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(3): 648-656.

图/表 12

图1

图2

图3

图4

表1

表2

表3

图5

图6

表4

表5

表6

参考文献 23

1	International Diabetes Federation. IDF diabetes atlas(9th edition)[EB/OL]. [2019-12-01].
2	Altaf F, Islam S M S, Akhtar N, et al. Going deep in medical image analysis: concepts, methods, challenges, and future directions[J]. IEEE Access, 2019, 7: 99540-99572.
3	Shen D G, Wu G R, Suk H. Deep learning in medical image analysis[J]. Annual Review of Biomedical Engineering, 2017, 19(1):221-248.
4	郜峰利, 陶敏, 李雪妍, 等. 基于深度学习的CT影像脑卒中精准分割[J]. 吉林大学学报: 工学版, 2020, 50(2): 678-684.
	Gao Feng-li, Tao Min, Li Xue-yan, et al. Accurate segmentation of stroke in CT image based on deep learning[J]. Journal of Jilin University (Engineering and Technology Edition), 2020, 50(2): 678-684.
5	陈雪云, 许韬, 黄小巧. 基于条件生成对抗网络的医学细胞图像生成检测方法[J].吉林大学学报:工学版, 2021, 51(4): 1414-1419.
	Chen Xue-yun, Xu Tao, Huang Xiao-qiao. Detection method of medical cell image generation based on conditional generative adversarial network[J]. Journal of Jilin University (Engineering and Technology Edition), 2021, 51(4): 1414-1419.
6	Gargeya R, Leng T. Automated identification of diabetic retinopathy using deep learning[J]. Ophthalmology, 2017, 124(7): 962-969.
7	Gulshan V, Peng L, Coram M, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs[J]. JAMA, 2016, 316(22): 2402-2410.
8	Gondal W M, Kohler J M, Grzeszick R, et al. Weakly-supervised localization of diabetic retinopathy lesions in retinal fundus images[C]∥IEEE International Conference on Image Processing, Beijing, China, 2017: 2069-2073.
9	Pratt H, Coenen F, Broadbent D M, et al. Convolutional neural networks for diabetic retinopathy[J]. Procedia Computer Science, 2016, 90: 200-205.
10	Wang Z, Yin Y, Shi J, et al. Zoom-in-net: deep mining lesions for diabetic retinopathy detection[C]∥International Conference on Medical Image Computing and Computer-Assisted Intervention, Cham, 2017: 267-275.
11	Bravo M A, Arbeláez P A. Automatic diabetic retinopathy classification[C]∥13th International Conference on Medical Information Processing and Analysis, International Society for Optics and Photonics, 2017: No.105721E.
12	Zhao Z Y, Zhang K R, Hao X J, et al. Bira-net: bilinear attention net for diabetic retinopathy grading[C]∥2019 IEEE International Conference on Image Processing, Taipei, China, 2019: 1385-1389.
13	Luo D, Kamata S I. Diabetic retinopathy grading based on lesion correlation graph[J/OL]. [2020-09-28].
14	Zhao Z Y, Chopra K, Zeng Z, et al. Sea-net: squeeze-and-excitation attention net for diabetic retinopathy grading[C]∥2020 IEEE International Conference on Image Processing, Abu Dhabi, United Arab Emirates, 2020: 2496-2500.
15	Zhou Y, He X D, Huang L, et al. Collaborative learning of semi-supervised segmentation and classification for medical images[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,Long Beach, CA, USA, 2019: 2074-2083.
16	Li X M, Hu X W, Yu L Q, et al. Canet: cross-disease attention network for joint diabetic retinopathy and diabetic macular edema grading[J]. IEEE Transactions on Medical Imaging, 2019, 39(5): 1483-1493.
17	Wang Z G, Yang J B. Diabetic retinopathy detection via deep convolutional networks for discriminative localization and visual explanation[J/OL]. [2020-09-29].
18	Jiang H Y, Yang K, Gao M D, et al. An interpretable ensemble deep learning model for diabetic retinopathy disease classification[C]∥2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Berlin, Germany, 2019: 2045-2048.
19	He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 2016: 770-778.
20	Woo S, Park J, Lee J Y, et al. Cbam: convolutional block attention module[C]∥Proceedings of the European Conference on Computer Vision, Munich, Germany, 2018: 3-19.
21	Kim Y, Park W, Roh M C, et al. Groupface: Learning latent groups and constructing group-based representations for face recognition[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA, 2020: 5621-5630.
22	Kaggle: diabetic retinopathy detection[EB/OL]. [2019-05-01].
23	Graham B. Kaggle diabetic retinopathy detection competition report[D]. Coventry, UK: University of Warwick, 2015.

相关文章 15

[1]	王雪,李占山,吕颖达. 基于多尺度感知和语义适配的医学图像分割算法[J]. 吉林大学学报(工学版), 2022, 52(3): 640-647.
[2]	康苏明,张叶娥. 基于Hadoop的跨社交网络局部时序链路预测算法[J]. 吉林大学学报(工学版), 2022, 52(3): 626-632.
[3]	宋林,王立平,吴军,关立文,刘知贵. 基于信息物理融合和数字孪生的可靠性分析[J]. 吉林大学学报(工学版), 2022, 52(2): 439-449.
[4]	曹洁,马佳林,黄黛麟,余萍. 一种基于多通道马尔可夫变迁场的故障诊断方法[J]. 吉林大学学报(工学版), 2022, 52(2): 491-496.
[5]	刘桂霞,裴志尧,宋佳智. 基于深度学习的蛋白质⁃ATP结合位点预测[J]. 吉林大学学报(工学版), 2022, 52(1): 187-194.
[6]	曲优,李文辉. 基于锚框变换的单阶段旋转目标检测方法[J]. 吉林大学学报(工学版), 2022, 52(1): 162-173.
[7]	赵宏伟,霍东升,王洁,李晓宁. 基于显著性检测的害虫图像分类[J]. 吉林大学学报(工学版), 2021, 51(6): 2174-2181.
[8]	张杰,景雯,陈富. 基于被动分簇算法的即时通信网络协议漏洞检测[J]. 吉林大学学报(工学版), 2021, 51(6): 2253-2258.
[9]	刘洲洲,张倩昀,马新华,彭寒. 基于优化离散差分进化算法的压缩感知信号重构[J]. 吉林大学学报(工学版), 2021, 51(6): 2246-2252.
[10]	王生生,陈境宇,卢奕南. 基于联邦学习和区块链的新冠肺炎胸部CT图像分割[J]. 吉林大学学报(工学版), 2021, 51(6): 2164-2173.
[11]	孙东明,胡亮,邢永恒,王峰. 基于文本融合的物联网触发动作编程模式服务推荐方法[J]. 吉林大学学报(工学版), 2021, 51(6): 2182-2189.
[12]	林俊聪,雷钧,陈萌,郭诗辉,高星,廖明宏. 基于电影视觉特性的动态多目标实时相机规划[J]. 吉林大学学报(工学版), 2021, 51(6): 2154-2163.
[13]	任丽莉,王志军,闫冬梅. 结合黏菌觅食行为的改进多元宇宙算法[J]. 吉林大学学报(工学版), 2021, 51(6): 2190-2197.
[14]	姚引娣,贺军瑾,李杨莉,谢荡远,李英. 自构建改进型鲸鱼优化BP神经网络的ET₀模拟计算[J]. 吉林大学学报(工学版), 2021, 51(5): 1798-1807.
[15]	董丽丽,杨丹,张翔. 基于深度学习的大规模语义文本重叠区域检索[J]. 吉林大学学报(工学版), 2021, 51(5): 1817-1822.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

分类方法	训练集数据量	实验结果
分类方法	训练集数据量	ACA	Macro-F1	Micro-F1
Pratt等^［9］	78 000	0.3410	0.3332	0.7376
Bravo等^［11］	127 480	0.5051	0.5081	0.5052
Zhao等^［12］	33 566	0.5431	0.5725	0.5436
Luo等^［13］	35 127	0.5496	0.5499	-
Zhao等^［14］	127 480	0.5994	0.6047	-
本文方法	7 500	0.6051	0.6079	0.6051

注意力机制	单分支分类器模型		双分支分类器模型
注意力机制	R-Net	RD-Net	BiR-Net	BiRD-Net
无注意力	0.5724	0.5839	0.5884	0.6032
有注意力	0.5744	0.5885	0.5910	0.6051

模型	R-Net	RA-Net	RD-Net	RAD-Net
单分支	0.8384	0.8423	-	-
双分支	0.8352	0.8416	0.8461	0.8558