GPT-4 Beats Eye Specialists In Retina & Glaucoma Management
A basic set of 20 questions (10 each for glaucoma and retina) from the American Academy of Ophthalmology¡¯s list of commonly asked questions by patients was randomly selected, along with 20 de-identified patient cases culled from Mount Sinai-affiliated eye clinics.
In the bustling corridors of New York's Mount Sinai Hospital, a breakthrough was unfolding in the field of ophthalmology. The air was charged with anticipation as researchers embarked on a journey to explore the symbiotic relationship between human expertise and artificial intelligence. At the helm of this exploration was GPT-4, a cutting-edge language model developed by OpenAI.
Lead author Andy Huang, a dedicated ophthalmology resident, described the discovery as "quite eye-opening." GPT-4, equipped with a vast reservoir of data, text, and images, emerged not just as an assistant but as a contender in the realm of glaucoma and retina disorders diagnosis and management.
The study, chronicled in JAMA Ophthalmology, pitted GPT-4 against a formidable lineup of 12 attending specialists and three seasoned trainees from the Department of Ophthalmology at the Icahn School of Medicine. A set of 20 questions, half dedicated to glaucoma and half to retina issues, served as the battleground. Additionally, 20 de-identified patient cases, drawn from Mount Sinai-affiliated eye clinics, were presented to both human specialists and the AI.
The revelation was stunning ¨C GPT-4 not only matched but, in certain instances, surpassed the accuracy and completeness of human assessments. In the realm of glaucoma, the AI demonstrated a superior performance, providing medical advice and case-management suggestions with remarkable precision. In the arena of retina questions, the balance tipped towards completeness, as the AI matched human accuracy and exceeded in providing comprehensive insights.
Louis R. Pasquale, Deputy Chair for Ophthalmology Research, expressed his surprise at the proficiency of AI in handling both glaucoma and retina cases. The implications were profound. Dr. Huang envisioned AI as a reliable assistant to eye specialists, offering diagnostic support and potentially lightening their workload, particularly in complex cases or high patient volume scenarios.
For patients, the integration of AI into mainstream ophthalmic practice heralded a new era. Quicker access to expert advice and more informed decision-making became tangible benefits. As the corridors of Mount Sinai echoed with the chatter of this groundbreaking research, the marriage of human expertise and artificial intelligence promised a brighter future for eye care.