Glaucoma Screening Using Artificial Intelligence Assisted Clinical Model in Singapore's Diabetic Eye Screening Program

NCT07243665 | Recruiting DMC

• 2 other identifiers: ECOS Ref: 2024-3461MOH-OFLCG21jun-0003
• Study Type: interventional
• Target: 1,040
• Locations: 1 country
• Sites: 1

Brief Summary

Glaucoma is major cause of irreversible blindness and is characterized by optic nerve damage and visual field loss. Screening for glaucoma is challenging due to lack of a simple, accurate, cost-efficient and standardized process. Artificial intelligence, (AI) especially deep learning (DL) algorithms have potential to automate glaucoma detection, but have to be evaluated in real world settings, before public deployment. This study aims to evaluate the screening accuracy of a DL algorithm for glaucoma detection using colour fundus photographs (CFP) in a pragmatic randomised control trial (RCT). The algorithm will be tested in 1040 eligible patients with diabetes, recruited from the Diabetes \& Metabolism Centre's clinics under the Singapore Integrated Diabetic Retinopathy Program (SiDRP) and randomized to 2 arms: AI-assisted model vs current standard of care (grader assessment). The performance of both arms will be compared to performance of study ophthalmologist in diagnosing glaucoma. We hypothesize that the DL model has better screening performance in detecting glaucoma in the community, compared to the current practice method.

Conditions

Glaucoma

Keywords

Glaucoma; deep learning; fundus photos; artificial intelligence; randomised controlled trial; screening

Outcome Measures

Primary Outcomes (1)

• Evaluation of model performance

  To compare the model performance in accuracy, sensitivity, specificity, positive predictive value and negative predictive value between the new AI-assisted clinical model and the current practice model in detecting glaucoma, with reference to the expert panel's standards.

  At study completion (after all fundus images have been graded and data collection is finalized; approximately within 12 months of study initiation)

Secondary Outcomes (2)

• Evaluation of time efficiency

  At study completion (after all fundus images have been graded and data collection is finalized; approximately within 12 months of study initiation)

• Evaluation of Grader's Acceptance

  At study completion (after all fundus images have been graded and data collection is finalized; approximately within 12 months of study initiation)

Study Arms (2)

Artificial Intelligence Assisted Arm

ACTIVE COMPARATOR

In this arm, human graders will review fundus photographs for glaucomatous features with the aid of output generated by an AI model trained to detect glaucoma. The AI output will be available during grading to support decision-making.

Diagnostic Test: Artificial Intelligence model to detect glaucoma

Current practice arm

PLACEBO COMPARATOR

Graders will assess fundus photographs for glaucoma following standard clinical practice, using a pre-specified and established set of diagnostic criteria without access to AI-generated outputs.

Other: No intervention

Interventions

Artificial Intelligence model to detect glaucoma DIAGNOSTIC_TEST

A Vision Transformer model to detect glaucoma from fundus photos

Also known as: Deep learning model, Vision Transformer, RetiGON

Artificial Intelligence Assisted Arm

No intervention OTHER

Control group with current practice model by human graders

Also known as: Control group, Current practice model

Current practice arm

Eligibility Criteria

• Age: 21 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)

You may qualify if:

• Aged 21 years old and above, with diabetes, including type 1 and type 2,
• Retinal photos of the patients can be taken with the fundus camera in the clinics, regardless of photos' quality, and
• They are willing and capable of providing a written informed consent form.

You may not qualify if:

• Patients who have difficulty in having retinal photos taken or have difficulties in completing the ocular examination protocols according to investigator's decision.
• Any other contraindication(s) as indicated by the endocrinologists responsible for the patients.

Sponsors & Collaborators

• Singapore Eye Research Institute: lead
• Singapore General Hospital: collaborator
• SingHealth Polyclinics: collaborator

Study Sites (1)

Singapore National Eye Centre

Singapore, Singapore, 168751, Singapore

RECRUITING

MeSH Terms

Conditions

Glaucoma

Interventions

Control Groups

Condition Hierarchy (Ancestors)

Ocular Hypertension; Eye Diseases

Intervention Hierarchy (Ancestors)

Epidemiologic Research Design; Epidemiologic Methods; Investigative Techniques; Research Design; Methods

Study Officials

• Ching-Yu Cheng, MD, PhD

  Singapore Eye Research Institute

  PRINCIPAL INVESTIGATOR

Central Study Contacts

Ching-Yu Cheng, MD, PhD

CONTACT

65767277; chingyu.cheng@duke-nus.edu.sg

Lavanya Raghavan, MD

CONTACT

65767201; raghavan.lavanya@seri.com.sg

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: SINGLE
• Who Masked: OUTCOMES ASSESSOR
• Purpose: DIAGNOSTIC
• Intervention Model: PARALLEL
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: November 16, 2025
• First Posted: November 24, 2025
• Study Start: November 17, 2025
• Primary Completion (Estimated): August 1, 2026
• Study Completion (Estimated): March 1, 2027
• Last Updated: January 29, 2026

Record last verified: 2026-01

Data Sharing

• IPD Sharing: Will share
• Shared Documents: SAP, CSR, ANALYTIC CODE
• Time Frame: 2028 onwards
• Access Criteria: Anonymised data only with the permission of the Principal Investigator

Locations

SG (1)