AI-Based Screening for Glaucoma

(GRaSS Trial)

Recruiting at 4 trial locations

Overseen By David S Friedman, MD, PhD, MPH

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: Massachusetts Eye and Ear Infirmary

Disqualifiers: Non-diabetic, others

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Compensation: Varies

Number of Visits: 2

2000 Participants Needed

What You Need to Know Before You Apply

What is the purpose of this trial?

This trial aims to determine if a new AI-based screening method can effectively detect glaucoma, a condition that can cause vision loss, during routine diabetic eye checks. AI will analyze participants' eye images, measure their eye pressure (IOP measurement), and possibly conduct a side vision test (Virtual Reality Visual Field Testing) if the AI detects potential glaucoma. This trial suits individuals with diabetes who already undergo eye screenings for diabetic retinopathy (a diabetes-related eye disease) at their primary care center. As an unphased trial, it offers participants the chance to contribute to cutting-edge research that could enhance early detection of glaucoma.

Will I have to stop taking my current medications?

The trial information does not specify whether you need to stop taking your current medications. It seems focused on eye screening, so it's unlikely to require changes to your medication, but you should confirm with the trial organizers.

What prior data suggests that this AI-based screening approach is safe for glaucoma detection?

Research has shown that using AI to screen for glaucoma is generally safe. Studies have found that AI can accurately detect signs of glaucoma from eye images, often outperforming human experts. This indicates that AI is reliable and poses no safety risks.

Checking eye pressure is also safe. It is a routine test for anyone over 40 during an eye exam, with no reports of harm from this process.

Using virtual reality (VR) to test vision is considered safe as well. It is reliable and easy to use, providing a safe method for checking side vision.

Overall, this AI-based method for glaucoma screening is well-tolerated and has demonstrated safety in other settings. ^{1 2 3 4 5}

Why are researchers excited about this trial?

Researchers are excited about the AI-based screening for glaucoma because it offers a new way to detect the disease early and efficiently. Unlike traditional methods that rely on manual examination and various tests, this AI system analyzes fundus photographs to spot signs of glaucoma and diabetic retinopathy simultaneously. If the AI flags potential glaucoma, virtual perimetry testing is used for further assessment, streamlining the process and potentially catching the disease sooner. This approach could make screenings faster, more accessible, and possibly more accurate, which is why it's generating interest.

What evidence suggests that this trial's AI-based glaucoma screening could be effective for detecting glaucoma?

This trial will compare AI-based glaucoma screening with standard diabetic eye disease (DED) screening. Studies have shown that AI can be very effective for glaucoma screening. In one study, the AI system correctly identified glaucoma in 88 to 90 percent of cases, while human experts were correct 79 to 81 percent of the time. Another study found that this AI method was highly accurate, correctly identifying people with glaucoma 95.4% of the time and those without it 87.3% of the time. Additionally, AI screening was 1.6 times more effective than traditional methods in spotting glaucoma, helping to reduce vision problems. These findings suggest that AI could greatly improve the accuracy and efficiency of glaucoma detection. Participants in this trial may receive AI-based screening as part of the experimental arm, which includes additional assessments if glaucoma is suspected. ^{1 3 6 7 8}

Who Is on the Research Team?

David S Friedman, MD, PhD, MPH

Principal Investigator

Massachusetts Eye and Ear

Are You a Good Fit for This Trial?

This trial is for individuals with diabetes who are getting screened for diabetic retinopathy using the LumineticsCore system. They must be able to give informed consent. People without diabetes or not undergoing this specific AI-based screening are excluded.

Inclusion Criteria

I have diabetes and am getting AI-based diabetic retinopathy screening.

I am willing and able to agree to participate in the study.

Timeline for a Trial Participant

Screening

Participants are screened for eligibility to participate in the trial

1 day

1 visit (in-person)

Screening and Assessment

Participants undergo fundus photography, intraocular pressure measurement, and virtual perimetry testing if needed

1 day

1 visit (in-person)

Follow-up

Participants are monitored for glaucoma detection and knowledge assessment

1 year

What Are the Treatments Tested in This Trial?

Interventions

• AI-based glaucoma screening
• IOP measurement
• Virtual Reality Visual Field Testing

Trial Overview The study tests a new glaucoma screening method that includes an AI algorithm analyzing eye photos, measuring eye pressure, and testing side vision with virtual reality if needed.

How Is the Trial Designed?

Group I: Combined DED and Glaucoma Screening Experimental Treatment 3 Interventions

Participants in this arm will have fundus photographs that are taken as a part of standard clinical care analyzed by AI for signs of glaucoma in addition to for diabetic retinopathy. They will also have intraocular pressure measured. If the AI detects possible glaucoma, participants will undergo virtual perimetry testing for further assessment.

Group II: DED Screening Only (Control Arm) Active Control 1 Intervention

Participants in this arm will undergo diabetic eye disease (DED) screening only for routine clinical care. This includes fundus photography for AI assessment of signs of diabetic retinopathy.

Who Is Running the Clinical Trial?

Massachusetts Eye and Ear Infirmary

Lead Sponsor

Trials

115

Recruited

15,000+

Centers for Disease Control and Prevention

Collaborator

Trials

902

Recruited

25,020,000+

Citations

1. aao.org aao.org/newsroom/news-releases/detail/artificial-intelligence-guided-glaucoma-screening-

Artificial Intelligence-Guided Glaucoma Screening Shows ... Results showed the algorithm correctly identified patients with glaucoma 88 to 90 percent of the time; human graders were correct 79 to 81 ...

2. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC9452706/

Glaucoma Screening: Is AI the Answer? - PMC A comprehensive screening program, if aided by AI, can mean significantly reduced costs and time, with significant improvements in yield and clinical outcomes.

3. nature.com nature.com/articles/s41746-025-01473-w

A hybrid multi model artificial intelligence approach for ... We present an AI-based Glaucoma Screening (AI-GS) network comprising six lightweight deep learning models (total size: 110 MB) that analyze fundus images.

4. glaucoma.org glaucoma.org/articles/easier-glaucoma-diagnosis-with-smartphone-apps-and-artificial-intelligence

Easier Glaucoma Diagnosis With Smartphone Apps And ... It achieved an area under the curve of 0.966 in clinical evaluations, with sensitivity and specificity rates of 95.4% and 87.3%, respectively.

5. valueinhealthjournal.com valueinhealthjournal.com/article/S1098-3015(25)02410-6/fulltext

The Cost-Effectiveness of an Artificial Intelligence-Based ... AI screening detected glaucoma 1.6 times more effectively than opportunistic case finding, reducing visual impairment by 0.7 months per ...

6. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC11946806/

Artificial Intelligence Models to Identify Patients at High Risk ... This study sought to develop artificial intelligence models that use self-reported health data from surveys to prescreen patients at high risk for glaucoma.

7. ophthalmologyscience.org ophthalmologyscience.org/article/S2666-9145(25)00049-1/fulltext

Comparison of Deep Learning and Clinician Performance ... Our findings highlight potential benefits of adopting AI-based strategies to improve the reproducibility, timeliness, and scalability of glaucoma care, which ...

8. sciencedirect.com sciencedirect.com/science/article/pii/S2666914524002070

Artificial Intelligence Models to Identify Patients with High ... Machine and deep learning models were able to use the extensive systematic data within EHR to identify individuals with glaucoma, without the need for ...