A deep learning system for myopia onset prediction and intervention effectiveness evaluation in children

Ziyi Qi; Tingyao Li; Jun Chen; Jason C Yam; Yang Wen; Gengyou Huang; Hua Zhong; Mingguang He; Dan Zhu; Rongping Dai; Bo Qian; Jingjing Wang; Chaoxu Qian; Wei Wang; Yanfei Zheng; Jian Zhang; Xianglong Yi; Zheyuan Wang; Bo Zhang; Chunyu Liu; Tianyu Cheng; Xiaokang Yang; Jun Li; Yan-Ting Pan; Xiaohu Ding; Ruilin Xiong; Yan Wang; Yan Zhou; Dagan Feng; Sichen Liu; Linlin Du; Jinliuxing Yang; Zhuoting Zhu; Lei Bi; Jinman Kim; Fangyao Tang; Yuzhou Zhang; Xiujuan Zhang; Haidong Zou; Marcus Ang; Clement C Tham; Carol Y Cheung; Chi Pui Pang; Bin Sheng; Xiangui He; Xun Xu

doi:10.1038/s41746-024-01204-7

A deep learning system for myopia onset prediction and intervention effectiveness evaluation in children

NPJ Digit Med. 2024 Aug 7;7(1):206. doi: 10.1038/s41746-024-01204-7.

Authors

Ziyi Qi^#^{1

2}, Tingyao Li^#^{3

4}, Jun Chen^#¹, Jason C Yam^#⁵, Yang Wen^#⁶, Gengyou Huang^{3

4}, Hua Zhong⁷, Mingguang He^{8

9}, Dan Zhu¹⁰, Rongping Dai¹¹, Bo Qian^{3

4}, Jingjing Wang¹, Chaoxu Qian⁷, Wei Wang⁹, Yanfei Zheng¹⁰, Jian Zhang¹², Xianglong Yi¹³, Zheyuan Wang^{3

4}, Bo Zhang¹, Chunyu Liu^{3

4}, Tianyu Cheng², Xiaokang Yang⁴, Jun Li¹⁴, Yan-Ting Pan¹⁴, Xiaohu Ding⁹, Ruilin Xiong⁹, Yan Wang¹⁰, Yan Zhou¹¹, Dagan Feng¹⁵, Sichen Liu¹, Linlin Du¹, Jinliuxing Yang¹, Zhuoting Zhu¹⁶, Lei Bi¹⁷, Jinman Kim¹⁵, Fangyao Tang⁵, Yuzhou Zhang⁵, Xiujuan Zhang⁵, Haidong Zou¹, Marcus Ang^{18

19}, Clement C Tham⁵, Carol Y Cheung⁵, Chi Pui Pang²⁰, Bin Sheng^{21

22}, Xiangui He^{23

24}, Xun Xu^{25

26}

Affiliations

¹ Department of Clinical Research, Shanghai Eye Disease Prevention and Treatment Center, Shanghai Eye Hospital, Shanghai Vision Health Center & Shanghai Children Myopia Institute, Shanghai, China.
² Department of Ophthalmology, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, National Clinical Research Center for Eye Diseases, Center of Eye Shanghai Key Laboratory of Ocular Fundus Diseases, Shanghai Engineering Center for Visual Science and Photomedicine, Shanghai, China.
³ Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, China.
⁴ MOE Key Laboratory of AI, School of Electronic, Information, and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, China.
⁵ Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, Kowloon, Hong Kong.
⁶ Guangdong Provincial Key Laboratory of Intelligent Information Processing, College of Electronic and Information Engineering, Shenzhen University, Shenzhen, China.
⁷ Department of Ophthalmology, First Affiliated Hospital of Kunming Medical University, Kunming, China.
⁸ Experimental Ophthalmology, The Hong Kong Polytechnic University, Kowloon, Hong Kong.
⁹ State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China.
¹⁰ Department of Ophthalmology, The Affiliated Hospital of Inner Mongolia Medical University, Hohhot, Inner Mongolia, China.
¹¹ Department of Ophthalmology, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, China.
¹² Department of Ophthalmology, Beijing Friendship Hospital Pinggu Campus, Capital Medical University, Beijing, China.
¹³ Department of Ophthalmology, The First Affiliated Hospital of Xinjiang Medical University, Urumqi, Xinjiang, China.
¹⁴ Affiliated Hospital of Yunnan University, Kunming, China.
¹⁵ School of Computer Science, The University of Sydney, Sydney, NSW, Australia.
¹⁶ Centre for Eye Research Australia, Ophthalmology, University of Melbourne, Melbourne, VIC, Australia.
¹⁷ Institute of Translational Medicine, Shanghai Jiao Tong University, Shanghai, China.
¹⁸ Singapore Eye Research Institute, Singapore National Eye Center, Singapore, Singapore.
¹⁹ DUKE-NUS Ophthalmology and Visual Sciences, Singapore, Singapore.
²⁰ Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, Kowloon, Hong Kong. cppang@cuhk.edu.hk.
²¹ Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, China. shengbin@sjtu.edu.cn.
²² MOE Key Laboratory of AI, School of Electronic, Information, and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, China. shengbin@sjtu.edu.cn.
²³ Department of Clinical Research, Shanghai Eye Disease Prevention and Treatment Center, Shanghai Eye Hospital, Shanghai Vision Health Center & Shanghai Children Myopia Institute, Shanghai, China. xianhezi@163.com.
²⁴ Department of Ophthalmology, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, National Clinical Research Center for Eye Diseases, Center of Eye Shanghai Key Laboratory of Ocular Fundus Diseases, Shanghai Engineering Center for Visual Science and Photomedicine, Shanghai, China. xianhezi@163.com.
²⁵ Department of Clinical Research, Shanghai Eye Disease Prevention and Treatment Center, Shanghai Eye Hospital, Shanghai Vision Health Center & Shanghai Children Myopia Institute, Shanghai, China. drxuxun@sjtu.edu.cn.
²⁶ Department of Ophthalmology, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, National Clinical Research Center for Eye Diseases, Center of Eye Shanghai Key Laboratory of Ocular Fundus Diseases, Shanghai Engineering Center for Visual Science and Photomedicine, Shanghai, China. drxuxun@sjtu.edu.cn.

^# Contributed equally.

Abstract

The increasing prevalence of myopia worldwide presents a significant public health challenge. A key strategy to combat myopia is with early detection and prediction in children as such examination allows for effective intervention using readily accessible imaging technique. To this end, we introduced DeepMyopia, an artificial intelligence (AI)-enabled decision support system to detect and predict myopia onset and facilitate targeted interventions for children at risk using routine retinal fundus images. Based on deep learning architecture, DeepMyopia had been trained and internally validated on a large cohort of retinal fundus images (n = 1,638,315) and then externally tested on datasets from seven sites in China (n = 22,060). Our results demonstrated robustness of DeepMyopia, with AUCs of 0.908, 0.813, and 0.810 for 1-, 2-, and 3-year myopia onset prediction with the internal test set, and AUCs of 0.796, 0.808, and 0.767 with the external test set. DeepMyopia also effectively stratified children into low- and high-risk groups (p < 0.001) in both test sets. In an emulated randomized controlled trial (eRCT) on the Shanghai outdoor cohort (n = 3303) where DeepMyopia showed effectiveness in myopia prevention compared to NonCyc-based model, with an adjusted relative reduction (ARR) of -17.8%, 95% CI: -29.4%, -6.4%. DeepMyopia-assisted interventions attained quality-adjusted life years (QALYs) of 0.75 (95% CI: 0.53, 1.04) per person and avoided blindness years of 13.54 (95% CI: 9.57, 18.83) per 1 million persons compared to natural lifestyle with no active intervention. Our findings demonstrated DeepMyopia as a reliable and efficient AI-based decision support system for intervention guidance for children.

Abstract

Grants and funding