Skip to main content

ORIGINAL RESEARCH article

Front. Cell Dev. Biol.
Sec. Molecular and Cellular Pathology
Volume 12 - 2024 | doi: 10.3389/fcell.2024.1517240
This article is part of the Research Topic Artificial Intelligence Applications in Chronic Ocular Diseases, Volume II View all 17 articles

Blinking characteristics analyzed by a deep learning model and its relationship with tear film stability in children with long-term use of orthokeratology

Provisionally accepted
Yue Wu Yue Wu 1Siyuan Wu Siyuan Wu 1Yinghai Yu Yinghai Yu 2,3Xiaojun Hu Xiaojun Hu 2Ting Zhao Ting Zhao 1*Yan Jiang Yan Jiang 1*Bilian Ke Bilian Ke 1*
  • 1 Department of Ophthalmology, Ren Ji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
  • 2 Department of Ophthalmology, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
  • 3 School of Health Science and Engineering, University of Shanghai for Science and Technology, Shanghai, China

The final, formatted version of the article will be published soon.

    Purpose: This study used a deep learning model to analyze blinking characteristics and their correlation with tear film stability in children using orthokeratology (ortho-k) lenses long-term. Methods: A retrospective case-control study included 31 children (58 eyes) with over a year of ortho-k use and 31 age- and gender-matched controls from 2021/09 to 2023/10. Comprehensive ophthalmological examinations were performed, including OSDI scoring, Keratograph 5M, and Lipiview. A deep learning system using U-Net and Swin-Transformer analyzed incomplete blinks (IB), complete blinks (CB), and the incomplete blinking rate (IBR) over 20 seconds. Relative IPH% indicated incomplete blinking extent. Model performance metrics and correlation analysis between blinking patterns and tear film stability were evaluated. Results: The deep learning system showed high accuracy (98.13%), precision (96.46%), sensitivity (98.10%), specificity (98.10%), and F1 score (0.9727). No significant differences were observed in OSDI scores, conjunctival redness, LLT, or tear meniscus height between groups. The ortho-k group had shorter non-invasive tear break-up times (NIBUT) and higher IB, IBR, relative IPH%, and prolonged eye-closing and opening phases compared to controls. Incomplete blinks negatively correlated with NIBUT. Conclusion: The deep learning system effectively assessed blinking characteristics. Children with long-term ortho-k use showed increased incomplete blinks and reduced tear film stability, emphasizing the need for monitoring in clinical follow-up.

    Keywords: Orthokeratology, Children, Blinking pattern, Tear film, Deep learning system

    Received: 25 Oct 2024; Accepted: 30 Dec 2024.

    Copyright: © 2024 Wu, Wu, Yu, Hu, Zhao, Jiang and Ke. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

    * Correspondence:
    Ting Zhao, Department of Ophthalmology, Ren Ji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Yan Jiang, Department of Ophthalmology, Ren Ji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Bilian Ke, Department of Ophthalmology, Ren Ji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China

    Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.