Development and Validation of a Deep Learning-based Myopia and Myopic Maculopathy Detection and Prediction System

Myopia has become a global public health issue. Myopia affects the psychological health of children and adolescents and poses a financial burden. Therefore, early detection and prediction of children at a high risk of myopia development and progression are critical for precise and effective interventions. In this study, we developed a deep learning system DeepMyopia, based on fundus images with the following objectives: 1) to predict myopia onset and progression; 2) To detect myopic macular degeneration for AI-assisted diagnosis; 3) To predict the development of myopic macular degeneration; 4) evaluate its cost-effectiveness.

Study Overview

• Status: Completed
• Conditions: Myopia; Myopic Macular Degeneration
• Intervention / Treatment: Diagnostic test: A deep learning-based myopia and myopic maculopathy detection and prediction system

Detailed Description

Myopia has become a global public health issue. Myopia affects the psychological health of children and adolescents and poses a financial burden. Furthermore, as myopia progresses it increases the risk of ocular complications such as myopic macular degeneration, leading to irreversible visual impairment or even blindness. According to the World Health Organization , more than 1 billion people worldwide are living with vision impairment caused by myopia, hyperopia, and other problems due to late detection. Therefore, early detection and prediction of children at a high risk of myopia development and progression are critical for precise and effective interventions.

In this study, we developed a deep learning system DeepMyopia, based on fundus images with the following objectives: 1) to predict myopia onset and progression; 2) To detect myopic macular degeneration for AI-assisted diagnosis; 3) To predict the development of myopic macular degeneration; 4) evaluate its cost-effectiveness.

• Study Type: Observational
• Enrollment (Actual): 30526

Contacts and Locations

Study Locations

• China
  • Shanghai
    • Shanghai, Shanghai, China, 200041
      • Shanghai Eye Disease Prevention and Treatment Center

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Probability Sample

Study Population

The SCALE, a prospective, school-based study, includes all children aged 4 to 14 years in Shanghai.

The SCALE-HM, a population-based, prospective, examiner-masked study, includes children and adolescents aged between 4 and 18 years with high myopia.

The STORM trial, a school-based, prospective, examiner-masked, cluster-randomized trial, includes children aged 6 to 9 years.

The SMS study is a school-based cross-sectional survey from Shanghai, including kindergarten and primary school students in Year 1 and 2.

The Beijing Children Eye study included children who came to the outpatient clinic of Beijing Friendship Hospital.

The JFFT study contains cross-sectional data from Shanghai, Yunnan, Inner Mongolia, Xinjiang and Guangzhou.

The Hong Kong Children Eye Study is a population-based cohort study of eye conditions in children aged 6-8 years.

Description

Inclusion Criteria:

1. Subjects with fundus images in the Shanghai Child and Adolescent Large-scale Eye Study (SCALE) ;
2. Subjects with fundus images in the Shanghai Time Outside to Reduce Myopia [STORM] trial;
3. Subjects with fundus images in the High Myopia Registration Study [SCALE-HM]
4. Subjects with fundus images in the Shanghai Myopia Screening (SMS) Study;
5. Subjects with fundus images in the Beijing Children Eye Study
6. Subjects with fundus images in the First Affiliated Hospital of Kunming Medical University;
7. Subjects with fundus images at the Ophthalmology Department of the First Affiliated Hospital of Xinjiang Medical University;
8. Subjects with fundus images at the Ophthalmology Department of the Affiliated Hospital of Inner Mongolia Medical University;
9. Subjects with fundus images at Zhongshan Eye Centre, Sun Yat-sen University;
10. Subjects with fundus images in the Hong Kong Children Eye Study;

Exclusion Criteria:

• Participants with poor-quality fundus images

Study Plan

How is the study designed?

Number of groups / cohorts

3

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
The training dataset; The training dataset was comprised of data from a school-based, prospective cohort (the Shanghai Time Outside to Reduce Myopia [STORM] trial) and data from another population-based, prospective study, the High Myopia Registration Study (SCALE-HM), with annual follow-up. Participants of the two studies were divided into a training set (70%), a tuning set (10%), and an internal test set (20%), which were not duplicated by each other at the participant level.	Diagnostic test: A deep learning-based myopia and myopic maculopathy detection and prediction system; This deep learning system is capable of analyzing fundus images for myopia staging, myopic maculopathy detection, cycloplegic refraction estimation and prediction, and risk stratification of myopia and myopic maculopathy onset.
The internal validation dataset; The internal validation dataset was comprised of data from a school-based, prospective cohort (the Shanghai Time Outside to Reduce Myopia [STORM] trial) and data from another population-based, prospective study, the High Myopia Registration Study (SCALE-HM), with annual follow-up. Participants of the two studies were divided into a training set (70%), a tuning set (10%), and an internal test set (20%), which were not duplicated by each other at the participant level.	Diagnostic test: A deep learning-based myopia and myopic maculopathy detection and prediction system; This deep learning system is capable of analyzing fundus images for myopia staging, myopic maculopathy detection, cycloplegic refraction estimation and prediction, and risk stratification of myopia and myopic maculopathy onset.
The external validation dataset; To test the extrapolation capabilities of the deep learning sysyem, two independent datasets, the Joint Five-site Fundus Test (JFFT) and the Hong Kong Children Eye Study (HKCES), were applied as external test sets. The JFFT study, a multi-site dataset, contains cross-sectional data from Shanghai, Yunnan, Inner Mongolia, Xinjiang and Guangzhou. HKCES, a population-based cohort study of eye conditions in children aged 6-8 years.	Diagnostic test: A deep learning-based myopia and myopic maculopathy detection and prediction system; This deep learning system is capable of analyzing fundus images for myopia staging, myopic maculopathy detection, cycloplegic refraction estimation and prediction, and risk stratification of myopia and myopic maculopathy onset.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
myopia staging detection possibility score	output of myopia staging task	immediately after inputting the data
myopic maculopathy detection possibility score	output of myopic maculopathy detection task	immediately after inputting the data
predicted spherical equivalent	output of assessing spherical equivalent task	immediately after inputting the data
predicted future annual spherical equivalent	output of predicting future spherical equivalent task	immediately after inputting the data
risk score of myopia and myopic maculopathy progression	output of the progression of myopia and myopic maculopathy predicion task	immediately after inputting the data

Collaborators and Investigators

• Sponsor: Shanghai Eye Disease Prevention and Treatment Center
• Collaborators: Beijing Friendship Hospital; Peking Union Medical College Hospital; Chinese University of Hong Kong; Zhongshan Ophthalmic Center, Sun Yat-sen University; Shanghai Jiao Tong University School of Medicine; First Affiliated Hospital of Xinjiang Medical University; The Affiliated Hospital of Inner Mongolia Medical University; First Affiliated Hospital of Kunming Medical University

Study record dates

Study Major Dates

• Study Start (Actual): April 1, 2022
• Primary Completion (Actual): April 1, 2023
• Study Completion (Actual): April 1, 2023

Study Registration Dates

• First Submitted: April 18, 2023
• First Submitted That Met QC Criteria: April 18, 2023
• First Posted (Actual): April 28, 2023

Study Record Updates

• Last Update Posted (Actual): April 28, 2023
• Last Update Submitted That Met QC Criteria: April 18, 2023
• Last Verified: April 1, 2023

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Eye Diseases; Retinal Degeneration; Retinal Diseases; Refractive Errors; Macular Degeneration; Myopia
• Other Study ID Numbers: 2022SQ023

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No