Related Experiment Video
Updated: Jul 2, 2025
03:14
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
558
Utility of artificial intelligence-based large language models in ophthalmic care.
Sayantan Biswas1, Leon N Davies1, Amy L Sheppard1
1School of Optometry, College of Health and Life Sciences, Aston University, Birmingham, UK.
Summary
Artificial intelligence (AI) large language models (LLMs) show promise in ophthalmology but require cautious adoption. Human judgment remains crucial for patient care and validating AI-generated information in eye health.
Area of Science:
- Ophthalmology
- Artificial Intelligence
- Medical Informatics
Background:
- Large Language Models (LLMs) like ChatGPT are increasingly utilized in scientific research.
- Their application and comparative performance in ophthalmic care are not yet fully understood.
- Existing studies highlight LLMs' utility in patient information, diagnosis, and examinations.
Purpose of the Study:
- To review and analyze the current applications of LLMs in ophthalmic care.
- To compare the performance of different LLMs against human experts in ophthalmology.
- To identify the potential and limitations of LLMs in the field of eye care.
Main Methods:
- Comprehensive literature review of recently published studies on LLMs in ophthalmology.
- Analysis of LLM performance metrics, including accuracy in diagnosis, symptom triaging, and examinations.
- Evaluation of factors influencing LLM performance, such as prompts and domain specificity.
Main Results:
- Human experts demonstrated highest proficiency (86%) in disease diagnosis.
- ChatGPT-4 excelled in symptom triaging (98%) and information provision (84.6%), and ophthalmology examinations (75.9%).
- LLMs showed superior performance in general ophthalmology but lower accuracy in subspecialties, with limitations including outdated training and hallucinations.
Conclusions:
- Ophthalmic professionals should adopt a conservative approach to AI, prioritizing human judgment in clinical decisions.
- Further exploration of LLM applications and potential in ophthalmology is needed.
- Establishing benchmarking standards and conducting clinical trials are crucial for the responsible deployment of LLMs in real-world ophthalmic settings.
