Using Machine Learning to Adapt Visual Aids for Patients With Low Vision

According to the WHO's definition of visual impairment, as of 2018, there were approximately 1.3 billion people with visual impairment in the world, and only 10% of countries can provide assisting services for the rehabilitation of visual impairment. Although China is one of the countries that can provide rehabilitation services for patients with visual impairment, due to restrictions on the number of professionals in various regions, uneven diagnosis and treatment, and regional differences in economic conditions, not all visually impaired patients can get the rehabilitation of assisting device fitting.

Traditional statistical methods were not enough to solve the problem of intelligent fitting of assisting devices. At present, there are almost no intelligent fitting models of assisting devices in the world. Therefore, in order to allow more low-vision patients to receive accurate and rapid rehabilitation services, we conducted a cross-sectional study on the assisting devices fitting for low-vision patients in Fujian Province, China in the past five years, and at the same time constructed a machine learning model to intelligently predict the adaptation result of the basic assisting devices for low vision patients.

Study Overview

• Status: Recruiting
• Conditions: Artificial Intelligence; Ophthalmology; Low Vision Aids
• Intervention / Treatment: Diagnostic test: Diagnostic test

• Study Type: Observational
• Enrollment (Anticipated): 400

Contacts and Locations

• Study Contact: Name: Jianmin Hu, M.D., Ph.D.; Phone Number: +8615359595888; Email: doctorhjm@163.com

Study Locations

• China
  • Fujian
    • Quanzhou, Fujian, China, 362000
      • Recruiting
      • 2nd Affilliated Hospital of Fujian Medical University
      • Contact: Jianmin Hu, M.D., Ph.D.; Phone Number: +8615359595888; Email: doctorhjm@163.com

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 3 years to 105 years (Child, Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

• Sampling Method: Non-Probability Sample

Study Population

Visually disabled patients were referred by the Town Disability Federation in Fujian Province and Guangdong Province

Description

Inclusion Criteria:

• Low vision
• Aged 3 to 105

Exclusion Criteria:

• Severe systemic disease
• Failure to sign informed consent or unwilling to participate

Study Plan

How is the study designed?

Number of groups / cohorts

3

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
Junior doctor group; Patients receive assisting devices fitting services from junior doctors	Diagnostic test: Diagnostic test; The training dataset was used to train the model, which was validated and tested by the other two datasets.
Senior doctor group; Patients receive assisting devices fitting services from senior doctors	Diagnostic test: Diagnostic test; The training dataset was used to train the model, which was validated and tested by the other two datasets.
Algorithm assisted group; Patients receive assisting devices fitting services from junior doctors assisted by the machine learning model	Diagnostic test: Diagnostic test; The training dataset was used to train the model, which was validated and tested by the other two datasets.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy of fitting results for assisting devices	The investigator will calculate the accuracy of fitting results for assisting devices in different group according to the ground truth.	baseline

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Time cost for fitting assisting devices	The investigator will calculate time cost for fitting assisting devices in different group.	baseline

Collaborators and Investigators

• Sponsor: Sun Yat-sen University

Study record dates

Study Major Dates

• Study Start (Actual): July 27, 2020
• Primary Completion (Anticipated): July 27, 2021
• Study Completion (Anticipated): December 27, 2021

Study Registration Dates

• First Submitted: May 17, 2021
• First Submitted That Met QC Criteria: May 18, 2021
• First Posted (Actual): May 19, 2021

Study Record Updates

• Last Update Posted (Actual): May 20, 2021
• Last Update Submitted That Met QC Criteria: May 19, 2021
• Last Verified: May 1, 2021

More Information

Terms related to this study

• Keywords: Artificial Intelligence; Low Vision
• Additional Relevant MeSH Terms: Nervous System Diseases; Eye Diseases; Neurologic Manifestations; Sensation Disorders; Vision Disorders; Vision, Low
• Other Study ID Numbers: SFLV-2020

Drug and device information, study documents

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