Early detection of visual impairment in young children using a smartphone-based deep learning system

Wenben Chen; Ruiyang Li; Qinji Yu; Andi Xu; Yile Feng; Ruixin Wang; Lanqin Zhao; Zhenzhe Lin; Yahan Yang; Duoru Lin; Xiaohang Wu; Jingjing Chen; Zhenzhen Liu; Yuxuan Wu; Kang Dang; Kexin Qiu; Zilong Wang; Ziheng Zhou; Dong Liu; Qianni Wu; Mingyuan Li; Yifan Xiang; Xiaoyan Li; Zhuoling Lin; Danqi Zeng; Yunjian Huang; Silang Mo; Xiucheng Huang; Shulin Sun; Jianmin Hu; Jun Zhao; Meirong Wei; Shoulong Hu; Liang Chen; Bingfa Dai; Huasheng Yang; Danping Huang; Xiaoming Lin; Lingyi Liang; Xiaoyan Ding; Yangfan Yang; Pengsen Wu; Feihui Zheng; Nick Stanojcic; Ji Peng Olivia Li; Carol Y. Cheung; Erping Long; Chuan Chen; Yi Zhu; Patrick Yu-Wai-Man; Ruixuan Wang; Wei shi Zheng; Xiaowei Ding; Haotian Lin

doi:10.1038/s41591-022-02180-9

Early detection of visual impairment in young children using a smartphone-based deep learning system

Wenben Chen, Ruiyang Li, Qinji Yu, Andi Xu, Yile Feng, Ruixin Wang, Lanqin Zhao, Zhenzhe Lin, Yahan Yang, Duoru Lin, Xiaohang Wu, Jingjing Chen, Zhenzhen Liu, Yuxuan Wu, Kang Dang, Kexin Qiu, Zilong Wang, Ziheng Zhou, Dong Liu, Qianni WuMingyuan Li, Yifan Xiang, Xiaoyan Li, Zhuoling Lin, Danqi Zeng, Yunjian Huang, Silang Mo, Xiucheng Huang, Shulin Sun, Jianmin Hu, Jun Zhao, Meirong Wei, Shoulong Hu, Liang Chen, Bingfa Dai, Huasheng Yang, Danping Huang, Xiaoming Lin, Lingyi Liang, Xiaoyan Ding, Yangfan Yang, Pengsen Wu, Feihui Zheng, Nick Stanojcic, Ji Peng Olivia Li, Carol Y. Cheung, Erping Long, Chuan Chen, Yi Zhu, Patrick Yu-Wai-Man, Ruixuan Wang, Wei shi Zheng, Xiaowei Ding^*, Haotian Lin^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

12 Citations (Scopus)

Abstract

Early detection of visual impairment is crucial but is frequently missed in young children, who are capable of only limited cooperation with standard vision tests. Although certain features of visually impaired children, such as facial appearance and ocular movements, can assist ophthalmic practice, applying these features to real-world screening remains challenging. Here, we present a mobile health (mHealth) system, the smartphone-based Apollo Infant Sight (AIS), which identifies visually impaired children with any of 16 ophthalmic disorders by recording and analyzing their gazing behaviors and facial features under visual stimuli. Videos from 3,652 children (≤48 months in age; 54.5% boys) were prospectively collected to develop and validate this system. For detecting visual impairment, AIS achieved an area under the receiver operating curve (AUC) of 0.940 in an internal validation set and an AUC of 0.843 in an external validation set collected in multiple ophthalmology clinics across China. In a further test of AIS for at-home implementation by untrained parents or caregivers using their smartphones, the system was able to adapt to different testing conditions and achieved an AUC of 0.859. This mHealth system has the potential to be used by healthcare professionals, parents and caregivers for identifying young children with visual impairment across a wide range of ophthalmic disorders.

Original language	English
Pages (from-to)	493-503
Number of pages	11
Journal	Nature Medicine
Volume	29
Issue number	2
DOIs	https://doi.org/10.1038/s41591-022-02180-9
Publication status	Published - Feb 2023
Externally published	Yes

Access to Document

10.1038/s41591-022-02180-9

Cite this

Chen, W., Li, R., Yu, Q., Xu, A., Feng, Y., Wang, R., Zhao, L., Lin, Z., Yang, Y., Lin, D., Wu, X., Chen, J., Liu, Z., Wu, Y., Dang, K., Qiu, K., Wang, Z., Zhou, Z., Liu, D., ... Lin, H. (2023). Early detection of visual impairment in young children using a smartphone-based deep learning system. Nature Medicine, 29(2), 493-503. https://doi.org/10.1038/s41591-022-02180-9

@article{b1b83a21c3ec492c941ed601049085e7,

title = "Early detection of visual impairment in young children using a smartphone-based deep learning system",

abstract = "Early detection of visual impairment is crucial but is frequently missed in young children, who are capable of only limited cooperation with standard vision tests. Although certain features of visually impaired children, such as facial appearance and ocular movements, can assist ophthalmic practice, applying these features to real-world screening remains challenging. Here, we present a mobile health (mHealth) system, the smartphone-based Apollo Infant Sight (AIS), which identifies visually impaired children with any of 16 ophthalmic disorders by recording and analyzing their gazing behaviors and facial features under visual stimuli. Videos from 3,652 children (≤48 months in age; 54.5% boys) were prospectively collected to develop and validate this system. For detecting visual impairment, AIS achieved an area under the receiver operating curve (AUC) of 0.940 in an internal validation set and an AUC of 0.843 in an external validation set collected in multiple ophthalmology clinics across China. In a further test of AIS for at-home implementation by untrained parents or caregivers using their smartphones, the system was able to adapt to different testing conditions and achieved an AUC of 0.859. This mHealth system has the potential to be used by healthcare professionals, parents and caregivers for identifying young children with visual impairment across a wide range of ophthalmic disorders.",

author = "Wenben Chen and Ruiyang Li and Qinji Yu and Andi Xu and Yile Feng and Ruixin Wang and Lanqin Zhao and Zhenzhe Lin and Yahan Yang and Duoru Lin and Xiaohang Wu and Jingjing Chen and Zhenzhen Liu and Yuxuan Wu and Kang Dang and Kexin Qiu and Zilong Wang and Ziheng Zhou and Dong Liu and Qianni Wu and Mingyuan Li and Yifan Xiang and Xiaoyan Li and Zhuoling Lin and Danqi Zeng and Yunjian Huang and Silang Mo and Xiucheng Huang and Shulin Sun and Jianmin Hu and Jun Zhao and Meirong Wei and Shoulong Hu and Liang Chen and Bingfa Dai and Huasheng Yang and Danping Huang and Xiaoming Lin and Lingyi Liang and Xiaoyan Ding and Yangfan Yang and Pengsen Wu and Feihui Zheng and Nick Stanojcic and Li, {Ji Peng Olivia} and Cheung, {Carol Y.} and Erping Long and Chuan Chen and Yi Zhu and Patrick Yu-Wai-Man and Ruixuan Wang and Zheng, {Wei shi} and Xiaowei Ding and Haotian Lin",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive licence to Springer Nature America, Inc.",

year = "2023",

month = feb,

doi = "10.1038/s41591-022-02180-9",

language = "English",

volume = "29",

pages = "493--503",

journal = "Nature Medicine",

issn = "1078-8956",

number = "2",

}

Chen, W, Li, R, Yu, Q, Xu, A, Feng, Y, Wang, R, Zhao, L, Lin, Z, Yang, Y, Lin, D, Wu, X, Chen, J, Liu, Z, Wu, Y, Dang, K, Qiu, K, Wang, Z, Zhou, Z, Liu, D, Wu, Q, Li, M, Xiang, Y, Li, X, Lin, Z, Zeng, D, Huang, Y, Mo, S, Huang, X, Sun, S, Hu, J, Zhao, J, Wei, M, Hu, S, Chen, L, Dai, B, Yang, H, Huang, D, Lin, X, Liang, L, Ding, X, Yang, Y, Wu, P, Zheng, F, Stanojcic, N, Li, JPO, Cheung, CY, Long, E, Chen, C, Zhu, Y, Yu-Wai-Man, P, Wang, R, Zheng, WS, Ding, X & Lin, H 2023, 'Early detection of visual impairment in young children using a smartphone-based deep learning system', Nature Medicine, vol. 29, no. 2, pp. 493-503. https://doi.org/10.1038/s41591-022-02180-9

TY - JOUR

T1 - Early detection of visual impairment in young children using a smartphone-based deep learning system

AU - Chen, Wenben

AU - Li, Ruiyang

AU - Yu, Qinji

AU - Xu, Andi

AU - Feng, Yile

AU - Wang, Ruixin

AU - Zhao, Lanqin

AU - Lin, Zhenzhe

AU - Yang, Yahan

AU - Lin, Duoru

AU - Wu, Xiaohang

AU - Chen, Jingjing

AU - Liu, Zhenzhen

AU - Wu, Yuxuan

AU - Dang, Kang

AU - Qiu, Kexin

AU - Wang, Zilong

AU - Zhou, Ziheng

AU - Liu, Dong

AU - Wu, Qianni

AU - Li, Mingyuan

AU - Xiang, Yifan

AU - Li, Xiaoyan

AU - Lin, Zhuoling

AU - Zeng, Danqi

AU - Huang, Yunjian

AU - Mo, Silang

AU - Huang, Xiucheng

AU - Sun, Shulin

AU - Hu, Jianmin

AU - Zhao, Jun

AU - Wei, Meirong

AU - Hu, Shoulong

AU - Chen, Liang

AU - Dai, Bingfa

AU - Yang, Huasheng

AU - Huang, Danping

AU - Lin, Xiaoming

AU - Liang, Lingyi

AU - Ding, Xiaoyan

AU - Yang, Yangfan

AU - Wu, Pengsen

AU - Zheng, Feihui

AU - Stanojcic, Nick

AU - Li, Ji Peng Olivia

AU - Cheung, Carol Y.

AU - Long, Erping

AU - Chen, Chuan

AU - Zhu, Yi

AU - Yu-Wai-Man, Patrick

AU - Wang, Ruixuan

AU - Zheng, Wei shi

AU - Ding, Xiaowei

AU - Lin, Haotian

PY - 2023/2

Y1 - 2023/2

N2 - Early detection of visual impairment is crucial but is frequently missed in young children, who are capable of only limited cooperation with standard vision tests. Although certain features of visually impaired children, such as facial appearance and ocular movements, can assist ophthalmic practice, applying these features to real-world screening remains challenging. Here, we present a mobile health (mHealth) system, the smartphone-based Apollo Infant Sight (AIS), which identifies visually impaired children with any of 16 ophthalmic disorders by recording and analyzing their gazing behaviors and facial features under visual stimuli. Videos from 3,652 children (≤48 months in age; 54.5% boys) were prospectively collected to develop and validate this system. For detecting visual impairment, AIS achieved an area under the receiver operating curve (AUC) of 0.940 in an internal validation set and an AUC of 0.843 in an external validation set collected in multiple ophthalmology clinics across China. In a further test of AIS for at-home implementation by untrained parents or caregivers using their smartphones, the system was able to adapt to different testing conditions and achieved an AUC of 0.859. This mHealth system has the potential to be used by healthcare professionals, parents and caregivers for identifying young children with visual impairment across a wide range of ophthalmic disorders.

AB - Early detection of visual impairment is crucial but is frequently missed in young children, who are capable of only limited cooperation with standard vision tests. Although certain features of visually impaired children, such as facial appearance and ocular movements, can assist ophthalmic practice, applying these features to real-world screening remains challenging. Here, we present a mobile health (mHealth) system, the smartphone-based Apollo Infant Sight (AIS), which identifies visually impaired children with any of 16 ophthalmic disorders by recording and analyzing their gazing behaviors and facial features under visual stimuli. Videos from 3,652 children (≤48 months in age; 54.5% boys) were prospectively collected to develop and validate this system. For detecting visual impairment, AIS achieved an area under the receiver operating curve (AUC) of 0.940 in an internal validation set and an AUC of 0.843 in an external validation set collected in multiple ophthalmology clinics across China. In a further test of AIS for at-home implementation by untrained parents or caregivers using their smartphones, the system was able to adapt to different testing conditions and achieved an AUC of 0.859. This mHealth system has the potential to be used by healthcare professionals, parents and caregivers for identifying young children with visual impairment across a wide range of ophthalmic disorders.

UR - http://www.scopus.com/inward/record.url?scp=85146859568&partnerID=8YFLogxK

U2 - 10.1038/s41591-022-02180-9

DO - 10.1038/s41591-022-02180-9

M3 - Article

C2 - 36702948

AN - SCOPUS:85146859568

SN - 1078-8956

VL - 29

SP - 493

EP - 503

JO - Nature Medicine

JF - Nature Medicine

IS - 2

ER -

Early detection of visual impairment in young children using a smartphone-based deep learning system

Abstract

Access to Document

Other files and links

Fingerprint

Cite this