AI Matches or Outperforms Ophthalmologists in Diagnosing Glaucoma, Retinal Diseases

AI Matches or Outperforms Ophthalmologists in Diagnosing Glaucoma, Retinal Diseases

February 23, 2024

A study conducted at the New York Eye and Ear Infirmary of Mount Sinai (NYEE) demonstrates that a large language model (LLM) AI system, specifically GPT-4 (Generative Pre-Training–Model 4) from OpenAI, can rival, and in some cases surpass, human ophthalmologists in diagnosing and treating patients with glaucoma and retina disease.

Unprecedented Performance of AI in Medical Diagnosis

Dr. Andy Huang, MD, an ophthalmology resident at NYEE and lead author of the study, expresses astonishment at the AI system's capabilities. "The performance of GPT-4 in our study was quite eye-opening," says Dr. Huang. "We recognized the enormous potential of this AI system from the moment we started testing it and were fascinated to observe that GPT-4 could not only assist but in some cases match or exceed the expertise of seasoned ophthalmic specialists."

Comparative Analysis: AI vs. Human Specialists

The research, led by a team of 12 attending specialists and three senior trainees from the Department of Ophthalmology at the Icahn School of Medicine, analyzed responses to a set of 20 questions and 20 deidentified patient cases related to glaucoma and retina conditions.

The findings reveal that the AI system demonstrated superior performance in response to glaucoma questions and case-management advice, while also achieving commendable accuracy and completeness in retina-related inquiries. Dr. Louis R. Pasquale, MD, FARVO, Deputy Chair for Ophthalmology Research for the Department of Ophthalmology and senior author of the study, emphasizes the potential impact of these results on clinical practice.

Transformative Potential of AI in Ophthalmology

Dr. Huang underscores the transformative potential of integrating AI into mainstream ophthalmic practice. "While further testing is necessary, the integration of AI could revolutionize patient care by providing quicker access to expert advice and informed decision-making," concluded Dr. Huang.

Reference

Assessment of a Large Language Model's Responses to Questions and Cases About Glaucoma and Retina Management, JAMA Ophthalmology (2024).