Detection of microaneurysms in fundus images based on improved YOLOv4 with SENet embedded_Journal of Biomedical Engineering

Authors：

 GAO Weiwei ¹ , SHAN Mingtao ¹ , SONG Nan ² , FAN Bo ¹ , FANG Yu ¹

1. Institute of Mechanical and Automotive Engineering, Shanghai University of Engineering Science, Shanghai 201620, P. R. China;
2. Eye & ENT Hospital of Fudan University, Shanghai 200031, P. R. China;

Corresponding author：

GAO Weiwei, Email: gww03020234@sina.com

Keywords：

Fundus image; You only look once v4; Squeeze-and-Excitation networks; Improved and fast fuzzy c-means; Microaneurysms

DOI：

10.7507/1001-5515.202203022

Video：

Export PDF Favorites Scan Get Citation

Abstract Full text Figures/Tables Video References Cited by

Microaneurysm is the initial symptom of diabetic retinopathy. Eliminating this lesion can effectively prevent diabetic retinopathy in the early stage. However, due to the complex retinal structure and the different brightness and contrast of fundus image because of different factors such as patients, environment and acquisition equipment, the existing detection algorithms are difficult to achieve the accurate detection and location of the lesion. Therefore, an improved detection algorithm of you only look once (YOLO) v4 with Squeeze-and-Excitation networks (SENet) embedded was proposed. Firstly, an improved and fast fuzzy c-means clustering algorithm was used to optimize the anchor parameters of the target samples to improve the matching degree between the anchors and the feature graphs; Then, the SENet attention module was embedded in the backbone network to enhance the key information of the image and suppress the background information of the image, so as to improve the confidence of microaneurysms; In addition, an spatial pyramid pooling was added to the network neck to enhance the acceptance domain of the output characteristics of the backbone network, so as to help separate important context information; Finally, the model was verified on the Kaggle diabetic retinopathy dataset and compared with other methods. The experimental results showed that compared with other YOLOv4 network models with various structures, the improved YOLOv4 network model could significantly improve the automatic detection results such as F-score which increased by 12.68%; Compared with other network models and methods, the automatic detection accuracy of the improved YOLOv4 network model with SENet embedded was obviously better, and accurate positioning could be realized. Therefore, the proposed YOLOv4 algorithm with SENet embedded has better performance, and can accurately and effectively detect and locate microaneurysms in fundus images.

Citation： GAO Weiwei, SHAN Mingtao, SONG Nan, FAN Bo, FANG Yu. Detection of microaneurysms in fundus images based on improved YOLOv4 with SENet embedded. Journal of Biomedical Engineering, 2022, 39(4): 713-720. doi: 10.7507/1001-5515.202203022 Copy

1.	Wong T Y, Sun J, Kawasaki R, et al. Guidelines on diabetic eye care: The International Council of Ophthalmology Recommendations for screening, follow-up, referral, and treatment based on resource setting. Ophthalmology, 2018, 125(10): 1608-1622.
2.	邵毅, 周召, 葛倩敏. 糖尿病视网膜病变及黄斑水肿诊疗规范: 英国皇家眼科医师学会指南解读. 眼科新进展, 2021, 41(7): 601-607.
3.	彭旭超, 赵艳莉, 林泰平, 等. 中老年糖尿病衰弱患者与肠道菌群结构及功能相关性研究. 生物医学工程学杂志, 2021, 38(6): 1126-1134.
4.	陈奕璇, 温力, 裴存文, 等. 非增殖期糖尿病视网膜病变微血管直径的变化. 国际眼科杂志, 2021, 21(9): 1632-1636.
5.	弓欣, 李淼. 光学相干断层扫描血管成像在全视网膜光凝术治疗糖尿病视网膜病变患者预后评估中的临床价值. 临床医学研究与实践, 2022, 7(8): 105-108.
6.	邬艳蓉, 夏桂媚, 高清月, 等. 人工智能在糖尿病视网膜病变筛查中的应用进展. 中华眼底病杂志, 2021, 37(6): 491-495.
7.	Zhang Xinpeng, Wu Jigang, Meng Min, et al. Feature-transfer network and local background suppression for microaneurysm detection. Mach Vision Appl, 2020, 32(1): 1-15.
8.	高玮玮, 沈建新, 王玉亮, 等. 免散瞳眼底图像中微动脉瘤的高效自动检测. 中国生物医学工程学报, 2013, 31(6): 839-845.
9.	Orlando J I, Prokofyeva E, Fresno M D, et al. An ensemble deep learning based approach for red lesion detection in fundus images. Comput Meth Prog Bio, 2017, 153: 115-127.
10.	Dashtbozorg B, Zhang J, Romeny B, et al. Retinal microaneurysms detection using local convergence index features. TIP, 2018, 39(2): 378-401.
11.	Xia H, Liao Y, Song S, et al. Microaneurysm detection in fundus images based on a novel end-to-end convolutional neural network. Biocybern Biomed Eng, 2021, 41(2): 589-604.
12.	赵学功, 邓佳坤, 魏浩然, 等. 基于卷积神经网络的眼底图像微血管瘤检测方法. 电子科技大学学报, 2021, 50(6): 915-920.
13.	Rohan R A. Film: finding the location of microaneurysms on the retina. Biomed Eng Lett, 2019, 9: 497-506.
14.	高玮玮, 沈建新, 王玉亮, 等. 改进的快速FCM及SVM实现糖网白色病灶的自动检测. 中国生物医学工程学报, 2013, 32(3): 305-312.
15.	范家伟, 张如如, 陆萌, 等. 深度学习方法在糖尿病视网膜病变诊断中的应用. 自动化学报, 2021, 47(5): 985-1004.
16.	Bochkovskiy A, Wang C, Liao H M. YOLOv4: optimal speed and accuracy of object detection. arXiv Preprint arXiv, 2020: 2004.10934.
17.	鞠默然, 罗海波, 王仲博, 等. 改进的YOLO V3算法及其在小目标检测中的应用. 光学学报, 2019, 39(7): 245-252.
18.	来文豪, 周孟然, 胡锋, 等. 基于多光谱成像和改进YOLOv4的煤矸石检测. 光学学报, 2020, 40(24): 1101-1109.
19.	Gao H, Zhuang L, Laurens M. Densely connected convolutional networks. arXiv Preprint arXiv, 2018: 1608. 069935v5.
20.	Yun S, Han D, Chun S, et al. CutMix: Regularization strategy to train strong classifiers with localizable features// International Conference on Computer Vision. Seoul: IEEE, 2019: 6023-6032.
21.	朱含露, 张旭中, 陈忻, 等. 基于横纵多尺度灰度差异加权双边滤波的弱小目标检测. 红外与毫米波学报, 2020, 39(4): 513-522.
22.	Ilya S, James M, George D, et al. On the importance of initialization and momentum in deep learning// Proceedings of the 30th International Conference on Machine Learning. Atlanta: IEEE, 2013: 1139-1147.
23.	Smith L N. Cyclical learning rates for training neural networks// 2017 IEEE Winter Conference on Applications of Computer Vision (WACV). Santa Rosa: IEEE, 2017: 401-410.
24.	李红宇, 黄志鹏, 张广玲, 等. 动态权重证据推理规则的CNN超参数质量评估. 小型微型计算机系统, 2021, 42(5): 1015-1021.
25.	秦风, 高原, 吴双. 基于改进贝叶斯非负Tikhonov正则化方法的同轴电缆信号传输畸变补偿研究. 电子与信息学报, 2021, 43(8): 2199-2206.
26.	濮玉, 朱俊江, 张德涛, 等. 基于改进卷积神经网络的房颤筛查算法. 生物医学工程学杂志, 2021, 38(4): 686-695.

1. Wong T Y, Sun J, Kawasaki R, et al. Guidelines on diabetic eye care: The International Council of Ophthalmology Recommendations for screening, follow-up, referral, and treatment based on resource setting. Ophthalmology, 2018, 125(10): 1608-1622.
2. 邵毅, 周召, 葛倩敏. 糖尿病视网膜病变及黄斑水肿诊疗规范: 英国皇家眼科医师学会指南解读. 眼科新进展, 2021, 41(7): 601-607.
3. 彭旭超, 赵艳莉, 林泰平, 等. 中老年糖尿病衰弱患者与肠道菌群结构及功能相关性研究. 生物医学工程学杂志, 2021, 38(6): 1126-1134.
4. 陈奕璇, 温力, 裴存文, 等. 非增殖期糖尿病视网膜病变微血管直径的变化. 国际眼科杂志, 2021, 21(9): 1632-1636.
5. 弓欣, 李淼. 光学相干断层扫描血管成像在全视网膜光凝术治疗糖尿病视网膜病变患者预后评估中的临床价值. 临床医学研究与实践, 2022, 7(8): 105-108.
6. 邬艳蓉, 夏桂媚, 高清月, 等. 人工智能在糖尿病视网膜病变筛查中的应用进展. 中华眼底病杂志, 2021, 37(6): 491-495.
7. Zhang Xinpeng, Wu Jigang, Meng Min, et al. Feature-transfer network and local background suppression for microaneurysm detection. Mach Vision Appl, 2020, 32(1): 1-15.
8. 高玮玮, 沈建新, 王玉亮, 等. 免散瞳眼底图像中微动脉瘤的高效自动检测. 中国生物医学工程学报, 2013, 31(6): 839-845.
9. Orlando J I, Prokofyeva E, Fresno M D, et al. An ensemble deep learning based approach for red lesion detection in fundus images. Comput Meth Prog Bio, 2017, 153: 115-127.
10. Dashtbozorg B, Zhang J, Romeny B, et al. Retinal microaneurysms detection using local convergence index features. TIP, 2018, 39(2): 378-401.
11. Xia H, Liao Y, Song S, et al. Microaneurysm detection in fundus images based on a novel end-to-end convolutional neural network. Biocybern Biomed Eng, 2021, 41(2): 589-604.
12. 赵学功, 邓佳坤, 魏浩然, 等. 基于卷积神经网络的眼底图像微血管瘤检测方法. 电子科技大学学报, 2021, 50(6): 915-920.
13. Rohan R A. Film: finding the location of microaneurysms on the retina. Biomed Eng Lett, 2019, 9: 497-506.
14. 高玮玮, 沈建新, 王玉亮, 等. 改进的快速FCM及SVM实现糖网白色病灶的自动检测. 中国生物医学工程学报, 2013, 32(3): 305-312.
15. 范家伟, 张如如, 陆萌, 等. 深度学习方法在糖尿病视网膜病变诊断中的应用. 自动化学报, 2021, 47(5): 985-1004.
16. Bochkovskiy A, Wang C, Liao H M. YOLOv4: optimal speed and accuracy of object detection. arXiv Preprint arXiv, 2020: 2004.10934.
17. 鞠默然, 罗海波, 王仲博, 等. 改进的YOLO V3算法及其在小目标检测中的应用. 光学学报, 2019, 39(7): 245-252.
18. 来文豪, 周孟然, 胡锋, 等. 基于多光谱成像和改进YOLOv4的煤矸石检测. 光学学报, 2020, 40(24): 1101-1109.
19. Gao H, Zhuang L, Laurens M. Densely connected convolutional networks. arXiv Preprint arXiv, 2018: 1608. 069935v5.
20. Yun S, Han D, Chun S, et al. CutMix: Regularization strategy to train strong classifiers with localizable features// International Conference on Computer Vision. Seoul: IEEE, 2019: 6023-6032.
21. 朱含露, 张旭中, 陈忻, 等. 基于横纵多尺度灰度差异加权双边滤波的弱小目标检测. 红外与毫米波学报, 2020, 39(4): 513-522.
22. Ilya S, James M, George D, et al. On the importance of initialization and momentum in deep learning// Proceedings of the 30th International Conference on Machine Learning. Atlanta: IEEE, 2013: 1139-1147.
23. Smith L N. Cyclical learning rates for training neural networks// 2017 IEEE Winter Conference on Applications of Computer Vision (WACV). Santa Rosa: IEEE, 2017: 401-410.
24. 李红宇, 黄志鹏, 张广玲, 等. 动态权重证据推理规则的CNN超参数质量评估. 小型微型计算机系统, 2021, 42(5): 1015-1021.
25. 秦风, 高原, 吴双. 基于改进贝叶斯非负Tikhonov正则化方法的同轴电缆信号传输畸变补偿研究. 电子与信息学报, 2021, 43(8): 2199-2206.
26. 濮玉, 朱俊江, 张德涛, 等. 基于改进卷积神经网络的房颤筛查算法. 生物医学工程学杂志, 2021, 38(4): 686-695.

Previous Article
ST segment morphological classification based on support vector machine multi feature fusion
Next Article
Lung parenchyma segmentation based on double scale parallel attention network

Journal of Biomedical Engineering

Detection of microaneurysms in fundus images based on improved YOLOv4 with SENet embedded

Abstract Full text Figures/Tables Video References Cited by

Previous Article

Next Article

Format

Content