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

Early detection of visual impairment in young children using a smartphone-based deep learning system. 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, J. O., Cheung, C. Y., Long, E., Chen, C., Zhu, Y., Yu-Wai-Man, P., Wang, R., Zheng, W., Ding, X., & Lin, H. Nature Medicine, 29(2):493–503, Nature Publishing Group, February, 2023. Number: 2

Paper doi abstract bibtex

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.

@article{chen_early_2023,
	title = {Early detection of visual impairment in young children using a smartphone-based deep learning system},
	volume = {29},
	copyright = {2023 The Author(s), under exclusive licence to Springer Nature America, Inc.},
	issn = {1546-170X},
	url = {https://www.nature.com/articles/s41591-022-02180-9},
	doi = {10.1038/s41591-022-02180-9},
	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.},
	language = {en},
	number = {2},
	urldate = {2023-05-16},
	journal = {Nature Medicine},
	publisher = {Nature Publishing Group},
	author = {Chen, Wenben and Li, Ruiyang and Yu, Qinji and Xu, Andi and Feng, Yile and Wang, Ruixin and Zhao, Lanqin and Lin, Zhenzhe and Yang, Yahan and Lin, Duoru and Wu, Xiaohang and Chen, Jingjing and Liu, Zhenzhen and Wu, Yuxuan and Dang, Kang and Qiu, Kexin and Wang, Zilong and Zhou, Ziheng and Liu, Dong and Wu, Qianni and Li, Mingyuan and Xiang, Yifan and Li, Xiaoyan and Lin, Zhuoling and Zeng, Danqi and Huang, Yunjian and Mo, Silang and Huang, Xiucheng and Sun, Shulin and Hu, Jianmin and Zhao, Jun and Wei, Meirong and Hu, Shoulong and Chen, Liang and Dai, Bingfa and Yang, Huasheng and Huang, Danping and Lin, Xiaoming and Liang, Lingyi and Ding, Xiaoyan and Yang, Yangfan and Wu, Pengsen and Zheng, Feihui and Stanojcic, Nick and Li, Ji-Peng Olivia and Cheung, Carol Y. and Long, Erping and Chen, Chuan and Zhu, Yi and Yu-Wai-Man, Patrick and Wang, Ruixuan and Zheng, Wei-shi and Ding, Xiaowei and Lin, Haotian},
	month = feb,
	year = {2023},
	note = {Number: 2},
	keywords = {Eye manifestations, Machine learning, Paediatrics, Translational research},
	pages = {493--503},
}

Downloads: 0