AI in Retina: Accuracy First

Ophthalmic patients want accuracy in their clinical decisions, regardless of who makes that decision

February 3, 2026
2 min

The Ophthalmologist

Overview

A study by the Macular Society reveals that patients with macular disease prioritize accuracy and verification over the identity of the decision-maker (AI vs. human) in retreatment decisions. The findings suggest that enhancing patient trust in AI requires a focus on performance metrics rather than the technology itself.

Background

As artificial intelligence (AI) tools transition from research to clinical practice in ophthalmology, understanding patient comfort with AI's role in treatment decisions is crucial. This is particularly relevant in macular services, where treatment choices can significantly impact vision outcomes. The study highlights the importance of patient preferences in the integration of AI into clinical workflows.

Data Highlights

Factor	Importance (%)
Error rate	34.4
Second reader/checker	33.6

Key Findings

Patients prioritize error rate and the presence of a second reader/checker in AI-led treatment decisions.
43% of participants had wet AMD, and 35% had dry AMD.
Participants showed no significant preference for human versus AI as the first reader.
Many patients suggested that a human making final decisions with AI support could be optimal.
Trust in AI can be enhanced by focusing on accuracy, transparency, and speed.

Clinical Implications

Clinicians should prioritize the accuracy and reliability of AI systems in macular disease management to foster patient trust. Emphasizing the role of human oversight in AI-assisted decisions may also improve patient acceptance and satisfaction.

Conclusion

The study underscores the need for a patient-centered approach in the adoption of AI in ophthalmology, focusing on performance and verification rather than the technology itself.