Embedded Files
Generative artificial intelligence matches eye doctors in terms of reasoning and recall
Generative artificial intelligence matches eye doctors in terms of reasoning and recall
Trialling state-of-the-art large language models (LLMs) against ophthalmologists at different stages of training sees the strongest LLM match expert consultant and attending ophthalmologists in answering challenging clinical questions.
University of Cambridge press release
Financial Times; BBC; Sky News; Der Spiegel; Evening Standard; MSN Italia; The National; Heraldo; Daily Mail; The Independent; Yahoo News
Optometry Today; National Health Executive; PharmaTimes; Science
Automating interpretation of vision tests can improve management of glaucoma
Automating interpretation of vision tests can improve management of glaucoma
Patients with blindness are eligible for various forms of social support, but require registration with local authorities which is often based on subjective clinical assessment. An automated algorithm for identifying blind glaucoma patients might improve the accessibility and fairness of registration processes and provision of support.
Cambridge University Hospitals press release
Wisbech Standard; Royston Crow; Oxford Eagle; Peterborough Matters
Healthcare Management; Eye News; Optometry Today; Optical News
Artificial intelligence and digital health interventions could transform eye care in lower income settings
Artificial intelligence and digital health interventions could transform eye care in lower income settings
Emerging technology could be used to improve eye health in lower income settings, particularly in regions where access to expertise is the limiting factor in providing safe, effective, and prompt care.
A new reporting guideline to improve chatbot health advice studies
A new reporting guideline to improve chatbot health advice studies
The Chatbot Assessment Reporting Tool (CHART) was developed by a large multinational and multidisciplinary team to improve the standard of research and reporting of studies evaluating health advice provided by chatbots such as ChatGPT, Gemini, Claude, and DeepSeek. CHART was co-published in JAMA Network Open, BJS, BMJ Medicine, Artificial Intelligence in Medicine, BMC Medicine, and Annals of Family Medicine, with an explanation and elaboration report published simultaneously in The BMJ.
Large language models drop performance in non-English languages
Large language models drop performance in non-English languages
Much is made of the medical potential of generative artificial intelligence, but recorded performance is lower in experiments involving non-English questions and answers. This has implications for equity of healthcare provision around the world as technology begins to be deployed in clinical settings.
The potential of general artificial intelligence in ophthalmology
The potential of general artificial intelligence in ophthalmology
Strengths and weaknesses of ChatGPT in general practice/primary care and ophthalmology
Strengths and weaknesses of ChatGPT in general practice/primary care and ophthalmology
Page updated
Google Sites
Report abuse