AI Screening for Diabetic Retinopathy

(DR-NeoRetina Trial)

Recruiting at 1 trial location

Overseen By Karim Hammamji, MD

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: Centre hospitalier de l'Université de Montréal (CHUM)

Disqualifiers: Under 18, Prior retinal treatment, others

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

Approved in 5 Jurisdictions This treatment is already approved in other countries

Compensation: Varies

Number of Visits: 1

630 Participants Needed

What You Need to Know Before You Apply

What is the purpose of this trial?

This trial tests NeoRetina, an AI tool designed to detect and assess the severity of diabetic retinopathy, an eye disease caused by diabetes. The trial aims to determine if this AI can match or outperform regular eye exams. Individuals with type 1 diabetes for at least five years or any type 2 diabetes, who receive care or are on a waiting list for eye evaluations at the Centre hospitalier de l'Université de Montréal, may qualify as candidates. The goal is to improve early detection and treatment planning for those at risk of vision problems due to diabetes. As an unphased trial, this study offers a unique opportunity to contribute to innovative research that could enhance future eye care for diabetic patients.

Will I have to stop taking my current medications?

The trial information does not specify whether you need to stop taking your current medications.

What prior data suggests that this AI algorithm is safe for screening diabetic retinopathy?

Research has shown that the NeoRetina AI tool is safe for checking diabetic retinopathy, an eye condition in people with diabetes. Studies have found this AI to be highly effective at examining eye images. Specifically, one study showed that NeoRetina correctly identified 91.4% of people with the disease and 95.4% of those without it.

Importantly, these studies did not report any negative effects from using the AI for screening. The process involves taking pictures of the eye, which is non-invasive and painless. Therefore, people using the NeoRetina AI can feel confident about its safety in detecting diabetes-related eye issues. ^{1 2 3 4 5}

Why are researchers excited about this trial?

Researchers are excited about the NeoRetina algorithm for diabetic retinopathy because it leverages artificial intelligence to enhance early detection of this eye condition. Traditional screenings rely on manual evaluations by ophthalmologists, which can be time-consuming and subject to human error. NeoRetina stands out by using AI to quickly analyze retinal images, potentially increasing accuracy and efficiency in diagnosing diabetic retinopathy. This technology could make screenings more accessible and faster, leading to earlier treatment and better outcomes for patients.

What evidence suggests that the NeoRetina algorithm is effective for detecting diabetic retinopathy?

Research has shown that artificial intelligence (AI) systems, such as NeoRetina, which is being tested in this trial, effectively spot diabetic retinopathy (DR), a condition affecting the eyes. Studies have demonstrated that AI can accurately examine eye images to detect signs of DR, often matching or even surpassing the accuracy of human specialists. One study found that AI systems excel at correctly identifying both those with the disease and those without it. This suggests that NeoRetina is likely very effective at detecting DR from eye images, which is crucial for initiating early treatment. Overall, AI's ability to detect DR appears promising for improving screening methods for this condition. ^{1 4 6 7 8}

Who Is on the Research Team?

Karim Hammamji, MD

Principal Investigator

Centre hospitalier de l'Université de Montréal (CHUM)

Are You a Good Fit for This Trial?

This trial is for adults over 18 with diabetes (Type 1 for at least 5 years, or Type 2) who are being treated or referred by the CHUM hospital. They must be able to give informed consent. It's not suitable for those who don't meet these specific conditions.

Inclusion Criteria

Ability to provide informed consent;

Diagnostic for diabetes : 3a) Type 1 diabetes of a lest 5 years of evolution; or 3b) Type 2 diabetes;

Diabetic patient followed and refered by a physician of the Centre hospitalier de l'Université de Montréal (CHUM) : 4a) followed by an endocrinologist of the CHUM; or 4b) hospitalized at the CHUM; or 4c) on the waiting list of the Ophthalmology Clinic of the CHUM for the evaluation of DR.

Timeline for a Trial Participant

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

AI Screening and Ophthalmological Evaluation

Participants undergo screening for diabetic retinopathy using the NeoRetina AI algorithm and a full eye examination by an ophthalmologist

Baseline

1 visit (in-person)

Follow-up

Participants are monitored for safety and effectiveness after the initial screening and evaluation

3 years

What Are the Treatments Tested in This Trial?

Interventions

• NeoRetina

Trial Overview The study tests NeoRetina, an AI algorithm designed to detect and grade diabetic retinopathy severity from eye photos. It compares routine eye exams and manual grading by ophthalmologists against this new AI screening method.

How Is the Trial Designed?

Group I: Diabetic Retinopathy (DR) Experimental Treatment 3 Interventions

Screening of DR with artificial intelligence (NeoRetina algorithm) and diagnostic evaluation with a standard of care ophthalmological examination.

NeoRetina is already approved in Canada, United States, European Union for the following indications:

Approved in Canada as NeoRetina for:

• Diabetic retinopathy screening

Approved in United States as CARA for:

• Visualization, storage, and enhancement of color fundus images

Approved in European Union as CARA for:

• Visualization, storage, and enhancement of color fundus images

Who Is Running the Clinical Trial?

Centre hospitalier de l'Université de Montréal (CHUM)

Lead Sponsor

Trials

389

Recruited

143,000+

DIAGNOS Inc.

Collaborator

Trials

1

Recruited

630+

Published Research Related to This Trial

A study involving 6146 patients showed that using AI for non-mydriatic fundus photography in diabetic retinopathy screening has a high sensitivity of 97.3% and a kappa coefficient of 0.75, indicating strong agreement with ophthalmic diagnoses.

The AI's consistency with fluorescein angiography was moderate, with a kappa coefficient of 0.53 and a coincidence rate of 66.9%, suggesting that AI can effectively assist in early detection of diabetic retinopathy, which is crucial for preventing blindness.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Application and observation of artificial intelligence in clinical practice of fundus screening for diabetic retinopathy with non-mydriatic fundus photography: a retrospective observational study of T2DM patients in Tianjin, China. Hao, Z., Xu, R., Huang, X., et al.[2022]

In a study analyzing 311,604 retinal images from 23,724 veterans, seven automated AI-based diabetic retinopathy (DR) screening algorithms were evaluated, revealing high negative predictive values (82.72-93.69%) but varying sensitivities (50.98-85.90%).

While some algorithms performed comparably or better than human graders, most did not exceed human performance, highlighting the need for thorough real-world testing of AI algorithms before they are used in clinical settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Multicenter, Head-to-Head, Real-World Validation Study of Seven Automated Artificial Intelligence Diabetic Retinopathy Screening Systems. Lee, AY., Yanagihara, RT., Lee, CS., et al.[2021]

The AI-based screening tool EyeWisdom®DSS demonstrated high sensitivity (91.0%) and good specificity (81.3%) for detecting diabetic retinopathy (DR) in a study of 549 type 2 diabetes patients, indicating its effectiveness as a screening method.

EyeWisdom®MCS was particularly effective at identifying patients without DR, with a specificity of 92.4%, suggesting that AI screening can be a valuable resource in areas with limited access to eye care.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Efficacy of artificial intelligence-based screening for diabetic retinopathy in type 2 diabetes mellitus patients. Pei, X., Yao, X., Yang, Y., et al.[2022]

In a study involving 600 diabetic patients, AI-based screening for diabetic retinopathy (DR) demonstrated high sensitivity (97.78%) and specificity (98.38%) for detecting referable DR, indicating its effectiveness as a diagnostic tool.

The use of AI not only matched but also improved the referral rate for DR cases compared to ophthalmologists, increasing referrals from 45 to 75, showcasing its potential to enhance clinical outcomes in community settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Application of artificial intelligence-based dual-modality analysis combining fundus photography and optical coherence tomography in diabetic retinopathy screening in a community hospital. Liu, R., Li, Q., Xu, F., et al.[2022]

An AI-based app was developed to predict retinal fluid in patients at risk for age-related macular degeneration (AMD) and diabetic retinopathy (DR), utilizing a convolutional neural network for accurate detection and monitoring.

The app demonstrated high accuracy in internal testing with an area under the curve (AUC) of 99.18% for fluid classification, indicating its potential effectiveness for early detection and personalized monitoring in clinical settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Predictive, preventive, and personalized management of retinal fluid via computer-aided detection app for optical coherence tomography scans. Quek, TC., Takahashi, K., Kang, HG., et al.[2023]

The AI algorithm Selena+ in the handheld Optomed Aurora fundus camera demonstrated a high sensitivity and specificity of 96.8% for detecting diabetic retinopathy (DR) in a study of 256 patients, making it a reliable screening tool.

The study found a perfect agreement (100%) between the AI analysis and the expert retina specialist's evaluation, suggesting that this technology could effectively support large-scale screening efforts for DR in real-world clinical settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Artificial intelligence in diabetic retinopathy screening: clinical assessment using handheld fundus camera in a real-life setting. Lupidi, M., Danieli, L., Fruttini, D., et al.[2023]

In a study involving 827 participants screened for diabetic retinopathy (DR) using AI, 33.2% were referred for follow-up, demonstrating the AI's effectiveness in identifying cases needing further evaluation.

The AI system showed a high sensitivity of 92% and specificity of 85% for detecting referable DR compared to human grading, indicating that AI can be a reliable tool for DR screening in clinical settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Feasibility and acceptance of artificial intelligence-based diabetic retinopathy screening in Rwanda. Whitestone, N., Nkurikiye, J., Patnaik, JL., et al.[2023]

The automated diabetic retinopathy screening tool, DART, demonstrated high sensitivity (94.6%) and a strong negative predictive value (98.1%) in analyzing ocular fundus photographs from 1123 diabetic eye exams, indicating its effectiveness in identifying potential cases of diabetic retinopathy.

With a specificity of 74.3% and an area under the ROC curve of 0.915, DART shows promise for implementation in Chile's national diabetic retinopathy screening program, suggesting it can be a reliable option in diverse healthcare settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Clinical validation of an artificial intelligence-based diabetic retinopathy screening tool for a national health system. Arenas-Cavalli, JT., Abarca, I., Rojas-Contreras, M., et al.[2023]

In a study of 241 patients using an AI system for diabetic retinal disease diagnosis, 51% had nondiagnostic results, with type 1 diabetes, smoking, and older age significantly increasing the likelihood of these results.

A new predictive dilation workflow was developed to identify patients who would benefit most from dilation before their examination, potentially improving efficiency and ensuring more at-risk patients receive necessary retinal screenings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Risk Factors for Nondiagnostic Imaging in a Real-World Deployment of Artificial Intelligence Diabetic Retinal Examinations in an Integrated Health care System: Maximizing Workflow Efficiency Through Predictive Dilation. Shou, BL., Venkatesh, K., Chen, C., et al.[2023]

The AIDRScreening system demonstrated high sensitivity (86.72%) and specificity (96.09%) in detecting referable diabetic retinopathy (RDR) among 962 participants with diabetes, indicating it is an effective tool for screening.

The system significantly improved diagnostic efficiency, achieving a 16.57% higher diagnostic accuracy compared to manual grading while reducing the average diagnostic time by 37.32%.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Performance of the AIDRScreening system in detecting diabetic retinopathy in the fundus photographs of Chinese patients: a prospective, multicenter, clinical study. Yang, Y., Pan, J., Yuan, M., et al.[2023]

The MONA.health artificial intelligence screening software demonstrated high diagnostic performance for detecting diabetic retinopathy (DR) and diabetic macular edema (DME), achieving an area under the curve (AUC) of 97.28% for DR and 98.08% for DME on a private test set.

The software maintained strong sensitivity (90.91%) and specificity (94.24%) across various subgroups, although sensitivity was slightly lower for individuals over 65 years old and Caucasians, indicating its effectiveness across diverse populations.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Artificial Intelligence Software for Diabetic Eye Screening: Diagnostic Performance and Impact of Stratification. Peeters, F., Rommes, S., Elen, B., et al.[2023]

Citations

1. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC11408885/

Implementation of Artificial Intelligence–Based Diabetic ... We evaluated the real-world performance of an artificial intelligence (AI) system that analyzes fundus images for DR screening in a Quebec tertiary care center.

2. journalretinavitreous.biomedcentral.com journalretinavitreous.biomedcentral.com/articles/10.1186/s40942-025-00670-9

The efficacy of artificial intelligence in diabetic retinopathy ... AI systems have demonstrated strong diagnostic performance in detecting diabetic retinopathy, with sensitivity and specificity comparable to or exceeding ...

3. ctv.veeva.com ctv.veeva.com/study/evaluation-of-neoretina-artificial-intelligence-algorithm-for-the-screening-of-diabetic-retinopathy

Evaluation of NeoRetina Artificial Intelligence Algorithm for the ... This prospective study aims to validate if NeoRetina, an artificial intelligence algorithm developped by DIAGNOS Inc. and trained to ...

4. sciencedirect.com sciencedirect.com/science/article/pii/S2352914824000017

Artificial intelligence for diabetic retinopathy detection This research article presents a novel method for DR detection, which is based on transfer learning to detect and classify DR lesions accurately.

5. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC12403315/

Diabetic retinopathy screening using machine learning Retinal fundus images are crucial for the diagnosis of various eye diseases, including diabetic retinopathy, glaucoma, and macular degeneration.

6. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC8546907/

Diabetic Retinopathy Screening with Automated Retinal ... By automated screening, 8.3% of the 180 study participants had referable diabetic eye disease, 13.3% had vision-threatening disease, and 29.4% had an ...

7. diabetesjournals.org diabetesjournals.org/care/article/46/10/1728/153626/Artificial-Intelligence-and-Diabetic-Retinopathy

Artificial Intelligence and Diabetic Retinopathy: AI Framework ... The algorithm had a sensitivity of 91.4% and a specificity of 95.4% for vtDR, which was superior to the performance of regional retina ...

8. nature.com nature.com/articles/s41467-023-44676-z

Autonomous artificial intelligence increases screening and ... We hypothesized that autonomous artificial intelligence (AI) diabetic eye exams at the point-of-care would increase diabetic eye exam completion rates.