[1]陈少滨,雷柏英,谢海,等.基于混合注意力的早产儿视网膜病变分类方法[J].深圳大学学报理工版,2022,39(6):701-708.[doi:10.3724/SP.J.1249.2022.06701]
 CHEN Shaobin,LEI Baiying,XIE Hai,et al.Classification of retinopathy of prematurity based on mixed attention[J].Journal of Shenzhen University Science and Engineering,2022,39(6):701-708.[doi:10.3724/SP.J.1249.2022.06701]
点击复制

基于混合注意力的早产儿视网膜病变分类方法()
分享到:

《深圳大学学报理工版》[ISSN:1000-2618/CN:44-1401/N]

卷:
第39卷
期数:
2022年第6期
页码:
701-708
栏目:
电子与信息科学
出版日期:
2022-11-15

文章信息/Info

Title:
Classification of retinopathy of prematurity based on mixed attention
文章编号:
202206012
作者:
陈少滨1雷柏英1谢海1张国明2杜曰山一2赵欣予2
1)深圳大学医学部,广东深圳 518071
2)深圳市眼科医院,暨南大学附属深圳眼科医院,深圳市眼病防治研究所,广东深圳 518040
Author(s):
CHEN Shaobin1 LEI Baiying1 XIE Hai1 ZHANG Guoming2 DU Yueshanyi2 and ZHAO Xinyu2
1) Faculty of Medicine, Shenzhen University, Shenzhen 518071, Guangdong Province, P.R.China
2) Shenzhen Eye Hospital, Jinan University, Shenzhen Eye Institute, Shenzhen 518040, Guangdong Province, P.R.China
关键词:
人工智能早产儿视网膜病变半监督学习自注意力空间注意力通道注意力
Keywords:
artificial intelligence retinopathy of prematurity semi-supervised learning self-attention spatial attention channel attention
分类号:
TP391
DOI:
10.3724/SP.J.1249.2022.06701
文献标志码:
A
摘要:
急性进展性后极部早产儿视网膜病变(aggressive posterior retinopathy of prematurity, AP-ROP)的患儿若不能及时得到诊断,易失去最佳治疗时间,导致患儿视网膜完全脱离,并最终失明.为更好地从ROP病例中诊断出AP-ROP,提出一种基于混合注意力网络(mixed attention network, MANet)的半监督学习方法.该方法分别通过低级数据增强器和高级数据增强器获得未标记训练数据的两种不同级别的增强数据.先利用低级数据增强器获得的增强数据生成未标记数据的伪标签,所得伪标签进一步被用作通过高级数据增强器获得的增强数据的训练标签.在设计的分类器模型中,将自注意力、空间注意力和通道注意力相结合,增强了分类器的特征提取能力.实验结果表明,本研究方法可以用很少的标记数据实现能够与全监督学习相媲美的性能,从而实现早产儿视网膜病变的自动检测,辅助医生进行ROP和AP-ROP的早期筛查.
Abstract:
Children with aggressive posterior retinopathy of prematurity (AP-ROP) are prone to lose the best treatment time if they are not diagnosed in time, leading to complete retinal detachment and eventual blindness. In order to diagnose this special type of AP-ROP from ROP cases more accurately, we propose a semi-supervised learning method based on mixed attention network. We obtain two different data representations of unlabeled training data through a weak enhancer and a strong enhancer, respectively. The enhanced data obtained by the weak enhancer is used to generate pseudo labels of the unlabeled data, which will be further used as the supervised training labels of enhanced data obtained by the strong enhancer. In the devised model, we consider the self-attention, spatial attention and channel attention together to enhance the capabilities of feature extraction. Experimental results show that the proposed method can achieve performance comparable to fully-supervised learning with few labeled data, thereby realizing automatic detection of retinopathy of prematurity, and assisting doctors in the early screening of ROP and AP-ROP.

参考文献/References:

[1] HELLSTR?M A, SMITH L E, DAMMANN O. Retinopathy of prematurity [J]. The Lancet, 2013, 382(9902): 1445-1457.
[2] 尹虹,黎晓新,李慧玲,等.早产儿视网膜病变的筛查及其相关因素分析[J].中华眼科杂志,2005,41(4):295-299.
YIN Hong, LI Xiaoxing, LI Huiling, et al. Screening of retinopathy of prematurity and analysis of its related factors [J]. Chinese Journal of Ophthalmology, 2005, 41(4): 295-299.(in Chinese)
[3] BLENCOWE H, LAWN J E, VAZQUEZ T, et al. Preterm-associated visual impairment and estimates of retinopathy of prematurity at regional and global levels for 2010 [J]. Pediatric Research, 2013, 74(Suppl.1): 35-49.
[4] 曹静,朱艳萍,李明霞.早产儿视网膜病变发病情况对比研究[J].中国新生儿科杂志,2016,31(5):330-334.
CAO Jing, ZHU Yanping, LI Mingxia. A comparative study on the incidence of retinopathy of prematurity [J]. Chinese Journal of Neonatology, 2016, 31(5): 330-334.(in Chinese)
[5] An International Committee for the Classification of Retinopathy of Prematurity. The international classification of retinopathy of prematurity revisited [J]. Archives of Ophthalmology, 2005, 123(7): 991-999.
[6] ZHANG Yinsheng, WANG Li, WU Zhenquan, et al. Development of an automated screening system for retinopathy of prematurity using a deep neural network for wide-angle retinal images [J]. IEEE Access, 2018: 7: 10232-10241.
[7] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Imagenet classification with deep convolutional neural networks [J]. Communications of the ACM, 2017,60(6): 84-90.
[8] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition [EB/OL].(2014-09-04) [2015-04-10].https://arxiv.org/pdf/1409.1556.pdf.
[9] WANG Jianyong, JU Rong, CHEN Yuanyuan, et al. Automated retinopathy of prematurity screening using deep neural networks [J]. EBioMedicine, 2018, 35: 361-368.
[10] HU Jie, SHEN Li, ALBANIE S, et al. Squeeze-and-excitation networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8): 2011-2023.
[11] 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: Springer, 2018: 3-19.
[12] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need [C]// Proceedings of the 31th International Conference on Neural Information Processing Systems. Red Hook, USA: Curran Associates Inc, 2017: 6000-6010.
[13] DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. An image is worth 16 × 16 words: transformers for image recognition at scale [EB/OL]. (2020-10-22)[2021-06-03]. https://arxiv.org/pdf/2010.11929.pdf
[14] SRINIVAS A, LIN T-Y, PARMAR N, et al. Bottleneck transformers for visual recognition [C]// Conference On Computer Vision and Pattern Recognition. Nashville, USA: IEEE, 2021: 16514-16524.
[15] CHAPELLE O, SCHOLKOPF B, ZIEN A. Semi-supervised learning (chapelle, o. et al., eds.; 2006) [book reviews] [J]. IEEE Transactions on Neural Networks, 2009, 20(3): 542-542.
[16] LEE D-H. Pseudo-label: the simple and efficient semi-supervised learning method for deep neural networks [EB/OL] // [2013-01-31]. https://www.kaggle.com/blobs/download/forum-message-attachment-files/746/pseudo_label_final.pdf
[17] SAJJADI M, JAVANMARDI M, TASDIZEN T. Regularization with stochastic transformations and perturbations for deep semi-supervised learning [C]// Proceedings of the 31th International Conference on Neural Information Processing Systems. Red Hook, USA: Curran Associates Inc, 2016: 1171-1179.
[18] TARVAINEN A, VALPOLA H. Mean teachers are better role models: weight-averaged consistency targets improve semi-supervised deep learning results [C]// Proceedings of the 31th International Conference on Neural Information Processing Systems. Red Hook, USA: Curran Associates Inc, 2017: 1195-1204.
[19] MIYATO T, MAEDA S-I, KOYAMA M, et al. Virtual adversarial training: a regularization method for supervised and semi-supervised learning [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 41(8): 1979-1993.
[20] GOODFELLOW I J, SHLENS J, SZEGEDY C. Explaining and harnessing adversarial examples [EB/OL]. (2014-10-20) [2015-02-25]. https://arxiv.org/pdf/1412.6572.pdf.
[21] SOHN K, BERTHELOT D, LI C-L, et al. Fixmatch: simplifying semi-supervised learning with consistency and confidence [C]// The 34th Conference on Neural Information Proceesing Systems. Red Hook, USA: Curran Associates Inc, 2020: 596-608.
[22] SHAW P, USZKOREIT J, VASWANI A. Self-attention with relative position representations [EB/OL]. (2018-03-06) [2018-04-12]. https://arxiv.org/pdf/1803.02155.pdf.
[23] SELVARAJU R R, COGSWELL M, DAS A, et al. Grad-CAM: visual explanations from deep networks via gradient-based localization [C]// Proceedings of the IEEE International Conference on Computer Vision. Venice, Italy: IEEE Computer Society, 2017: 618-626.

相似文献/References:

[1]潘长城,徐晨,李国.解全局优化问题的差分进化策略[J].深圳大学学报理工版,2008,25(2):211.
 PAN Chang-cheng,XU Chen,and LI Guo.Differential evolutionary strategies for global optimization[J].Journal of Shenzhen University Science and Engineering,2008,25(6):211.
[2]骆剑平,李霞.求解TSP的改进混合蛙跳算法[J].深圳大学学报理工版,2010,27(2):173.
 LUO Jian-ping and LI Xia.Improved shuffled frog leaping algorithm for solving TSP[J].Journal of Shenzhen University Science and Engineering,2010,27(6):173.
[3]蔡良伟,李霞.基于混合蛙跳算法的作业车间调度优化[J].深圳大学学报理工版,2010,27(4):391.
 CAI Liang-wei and LI Xia.Optimization of job shop scheduling based on shuffled frog leaping algorithm[J].Journal of Shenzhen University Science and Engineering,2010,27(6):391.
[4]张重毅,刘彦斌,于繁华,等.CDA市场环境模型进化研究[J].深圳大学学报理工版,2010,27(4):413.
 ZHANG Zhong-yi,LIU Yan-bin,YU Fan-hua,et al.Research on the evolution model of CDA market environment[J].Journal of Shenzhen University Science and Engineering,2010,27(6):413.
[5]姜建国,周佳薇,郑迎春,等.一种双菌群细菌觅食优化算法[J].深圳大学学报理工版,2014,31(1):43.[doi:10.3724/SP.J.1249.2014.01043]
 Jiang Jianguo,Zhou Jiawei,Zheng Yingchun,et al.A double flora bacteria foraging optimization algorithm[J].Journal of Shenzhen University Science and Engineering,2014,31(6):43.[doi:10.3724/SP.J.1249.2014.01043]
[6]蔡良伟,刘思麒,李霞,等.基于蚁群优化的正则表达式分组算法[J].深圳大学学报理工版,2014,31(3):279.[doi:10.3724/SP.J.1249.2014.03279]
 Cai Liangwei,Liu Siqi,Li Xia,et al.Regular expression grouping algorithm based on ant colony optimization[J].Journal of Shenzhen University Science and Engineering,2014,31(6):279.[doi:10.3724/SP.J.1249.2014.03279]
[7]宁剑平,王冰,李洪儒,等.递减步长果蝇优化算法及应用[J].深圳大学学报理工版,2014,31(4):367.[doi:10.3724/SP.J.1249.2014.04367]
 Ning Jianping,Wang Bing,Li Hongru,et al.Research on and application of diminishing step fruit fly optimization algorithm[J].Journal of Shenzhen University Science and Engineering,2014,31(6):367.[doi:10.3724/SP.J.1249.2014.04367]
[8]刘万峰,李霞.车辆路径问题的快速多邻域迭代局部搜索算法[J].深圳大学学报理工版,2015,32(2):196.[doi:10.3724/SP.J.1249.2015.02000]
 Liu Wanfeng,and Li Xia,A fast multi-neighborhood iterated local search algorithm for vehicle routing problem[J].Journal of Shenzhen University Science and Engineering,2015,32(6):196.[doi:10.3724/SP.J.1249.2015.02000]
[9]蔡良伟,程璐,李军,等.基于遗传算法的正则表达式规则分组优化[J].深圳大学学报理工版,2015,32(3):281.[doi:10.3724/SP.J.1249.2015.03281]
 Cai Liangwei,Cheng Lu,Li Jun,et al.Regular expression grouping optimization based on genetic algorithm[J].Journal of Shenzhen University Science and Engineering,2015,32(6):281.[doi:10.3724/SP.J.1249.2015.03281]
[10]王守觉,鲁华祥,陈向东,等.人工神经网络硬件化途径与神经计算机研究[J].深圳大学学报理工版,1997,14(1):8.
 Wang Shoujue,Lu Huaxiang,Chen Xiangdong and Zeng Yujuan.On the Hardware for Artificial Neural Networks and Neurocomputer[J].Journal of Shenzhen University Science and Engineering,1997,14(6):8.

备注/Memo

备注/Memo:
Received: 2022- 02-16; Accepted: 2022-08-25; Online (CNKI): 2022-10-20
Foundation: National Natural Science Foundation of China (61871274, 62106153)
Corresponding author: Professor LEI Baiying.E-mail: leiby@szu.edu.cn
Citation: CHEN Shaobin, LEI Baiying, XIE Hai, et al. Classification of retinopathy of prematurity based on mixed attention [J]. Journal of Shenzhen University Science and Engineering, 2022, 39(6): 701-708.(in Chinese)
基金项目:国家自然科学基金资助项目(61871274,62106153)
作者简介:陈少滨(1996—),深圳大学硕士研究生.研究方向:医学图像处理.E-mail: 13531194616@163.com
引文:陈少滨,雷柏英,谢海,等.基于混合注意力的早产儿视网膜病变分类方法[J].深圳大学学报理工版,2022,39(6):701-708.
更新日期/Last Update: 2022-11-30