In an impressive development, the GPT-4, an artificial intelligence (AI) model by OpenAI, managed to prevail in diagnostic accuracy of a group of ophthalmologists, leaving the medical experts speechless. The study, published in the journal PLOS Digital Health, showed that the GPT-4 correctly answered 60 of 87 multiple-choice questions (MCQs) about ophthalmology, outperforming junior doctors and matching the results of more experienced ophthalmologists.
👁️ #HAVE is much better than non-specialist doctors at assessing eye problems and providing advice, Cambridge researchers have found.
GPT-4 could triage patients and decide which #ophthalmology issues are emergencies that need immediate attention👇 https://t.co/nX9OYQb1XR
— Cambridge University (@Cambridge_Uni) April 18, 2024
Notable Results:
- The GPT-4 scored 60 correct answers, compared to an average of 37 for younger doctors.
- Ophthalmologists in training reached 59,7 correct answers, while experienced ophthalmologists (except one) averaged 66,4.
- The study used a manual-based questionnaire to screen trainees, with content not widely available.
- GPT-4, equipped with GPT-4 or GPT-3.5, received three attempts for each question.
Importance for the Health Sector:
These results highlight the enormous potential of AI in the medical field. The GPT-4's ability to rival experienced ophthalmologists in diagnostic accuracy paves the way for a multitude of applications, including:
- Improving access to health care, especially in areas with a shortage of qualified doctors.
- Additional support to doctors for faster and more accurate diagnoses.
- Evaluating and analyzing medical data at scale.
Next steps:
While the study is an important milestone, the development of AI in the medical field must be done with caution and ethical responsibility. Continuous evaluation of data, ensuring transparency and developing ethical guidelines are essential for the responsible adoption of AI in healthcare.
Conclusion:
GPT-4's victory over ophthalmologists is an impressive achievement that highlights the enormous potential of AI in the medical field. With careful development and responsible adoption, TN has the potential to shape the future of healthcare for the benefit of patients around the world.
