Glaucoma Screening With Artificial Intelligence

The University of Hong Kong (HKU)

Status: Enrolling

Conditions: Glaucoma

Treatments: Diagnostic Test: Optic disc assessment by AI; Diagnostic Test: ROTA assessment by AI

Study type: Interventional

Funder types: Other

Identifiers

NCT06012058

H012_Protocol_Glaucoma

Details and patient eligibility

About

This randomized clinical trial aims to compare the diagnostic performance of two AI-enabled screening strategies - ROTA (RNFL optical texture analysis) assessment versus optic disc photography - in detecting glaucoma within a population-based sample. Secondary objectives are to (1) compare the diagnostic performance of ROTA AI assessment versus OCT RNFL thickness assessment by AI, and ROTA AI assessment versus OCT RNFL thickness assessment by trained graders, (2) investigate the cost-effectiveness of AI ROTA assessment for glaucoma screening, and (3) estimate the prevalence of glaucoma in Hong Kong.

Full description

Glaucoma is the leading cause of irreversible blindness affecting 76 million patients worldwide in 2020. Characterized by progressive degeneration of the optic nerve, early detection of disease deterioration with timely intervention is critical to prevent progressive loss in vision. In the 5th World Glaucoma Association Consensus Meeting, a diverse and representative group of glaucoma clinicians and scientists deliberated on the value and methods of glaucoma screening. Whereas it has been recognized that early detection of glaucoma for treatment is beneficial to preserve the quality of vision and quality of life as glaucoma treatments are often effective, easy to use and well tolerated, the optimal screening strategy for glaucoma has not yet been determined.

ROTA (Retinal Nerve Fiber Layer Optical Texture Analysis) is a patented algorithm designed to detect axonal fiber bundle loss in glaucoma. Unlike conventional Optical Coherence Tomography (OCT) analysis, ROTA uses non-linear transformation to reveal the optical textures and trajectories of axonal fiber bundles, allowing for intuitive and reliable recognition of RNFL abnormalities without the need for normative databases. It can be applied across different OCT models and is particularly effective at detecting focal RNFL defects in early glaucoma and varying degrees of RNFL damage in end-stage glaucoma. The proposed study will address whether the application AI on ROTA is feasible and cost-effective in the setting of glaucoma screening, and whether ROTA would outperform optic disc photography and OCT RNFL thickness assessment.

Enrollment: 3,175 estimated patients

Sex: All

Ages: 50+ years old

Volunteers: Accepts Healthy Volunteers

Inclusion criteria

• Individuals aged 50 years or above

Exclusion criteria

• Physically incapacitated
• Not able to cooperate for clinical examination or optical coherence tomography (OCT) investigation will be excluded

Trial design

Primary purpose: Screening

Allocation: Randomized

Interventional model: Parallel Assignment

Masking: None (Open label)

---

3,175 participants in 2 patient groups

Retinal nerve fiber layer optical texture analysis (ROTA)

Experimental group

Description:

The RNFL is imaged with OCT for ROTA.

Treatment:

Diagnostic Test: ROTA assessment by AI

Optic disc photography

Active Comparator group

Description:

The optic disc is imaged with color fundus camera.

Treatment:

Diagnostic Test: Optic disc assessment by AI