(Press-News.org) The clinical knowledge and reasoning skills of GPT-4 are approaching the level of specialist eye doctors, a study led by the University of Cambridge has found.
GPT-4 - a ‘large language model’ - was tested against doctors at different stages in their careers, including unspecialised junior doctors, and trainee and expert eye doctors. Each was presented with a series of 87 patient scenarios involving a specific eye problem, and asked to give a diagnosis or advise on treatment by selecting from four options.
GPT-4 scored significantly better in the test than unspecialised junior doctors, who are comparable to general practitioners in their level of specialist eye knowledge.
GPT-4 gained similar scores to trainee and expert eye doctors - although the top performing doctors scored higher.
The researchers say that large language models aren’t likely to replace healthcare professionals, but have the potential to improve healthcare as part of the clinical workflow.
They say state-of-the-art large language models like GPT-4 could be useful for providing eye-related advice, diagnosis, and management suggestions in well-controlled contexts, like triaging patients, or where access to specialist healthcare professionals is limited.
“We could realistically deploy AI in triaging patients with eye issues to decide which cases are emergencies that need to be seen by a specialist immediately, which can be seen by a GP, and which don’t need treatment,” said Dr Arun Thirunavukarasu, lead author of the study, which he carried out while a student at the University of Cambridge’s School of Clinical Medicine
He added: “The models could follow clear algorithms already in use, and we’ve found that GPT-4 is as good as expert clinicians at processing eye symptoms and signs to answer more complicated questions.
“With further development, large language models could also advise GPs who are struggling to get prompt advice from eye doctors. People in the UK are waiting longer than ever for eye care.
Large volumes of clinical text are needed to help fine-tune and develop these models, and work is ongoing around the world to facilitate this.
The researchers say that their study is superior to similar, previous studies because they compared the abilities of AI to practicing doctors, rather than to sets of examination results.
“Doctors aren't revising for exams for their whole career. We wanted to see how AI fared when pitted against to the on-the-spot knowledge and abilities of practicing doctors, to provide a fair comparison,” said Thirunavukarasu, who is now an Academic Foundation Doctor at Oxford University Hospitals NHS Foundation Trust.
He added: “We also need to characterise the capabilities and limitations of commercially available models, as patients may already be using them - rather than the internet - for advice.”
The test included questions about a huge range of eye problems, including extreme light sensitivity, decreased vision, lesions, itchy and painful eyes, taken from a textbook used to test trainee eye doctors. This textbook is not freely available on the internet, making it unlikely that its content was included in GPT-4’s training datasets.
The results are published today in the journal PLOS Digital Health.
“Even taking the future use of AI into account, I think doctors will continue to be in charge of patient care. The most important thing is to empower patients to decide whether they want computer systems to be involved or not. That will be an individual decision for each patient to make,” said Thirunavukarasu.
GPT-4 and GPT-3.5 – or ‘Generative Pre-trained Transformers’ - are trained on datasets containing hundreds of billions of words from articles, books, and other internet sources. These are two examples of large language models; others in wide use include Pathways Language Model 2 (PaLM 2) and Large Language Model Meta AI 2 (LLaMA 2).
The study also tested GPT-3.5, PaLM2, and LLaMA with the same set of questions. GPT-4 gave more accurate responses than all of them.
GPT-4 powers the online chatbot ChatGPT to provide bespoke responses to human queries. In recent months, ChatGPT has attracted significant attention in medicine for attaining passing level performance in medical school examinations, and providing more accurate and empathetic messages than human doctors in response to patient queries.
The field of artificially intelligent large language models is moving very rapidly. Since the study was conducted, more advanced models have been released - which may be even closer to the level of expert eye doctors.
END
Artificial Intelligence beats doctors in accurately assessing eye problems
2024-04-17
[Attachments] See images for this press release:
ELSE PRESS RELEASES FROM THIS DATE:
Ice age climate analysis reduces worst-case warming expected from rising CO2
2024-04-17
As carbon dioxide accumulates in the atmosphere, the Earth will get hotter. But exactly how much warming will result from a certain increase in CO2 is under study. The relationship between CO2 and warming, known as climate sensitivity, determines what future we should expect as CO2 levels continue to climb.
New research led by the University of Washington analyzes the most recent ice age, when a large swath of North America was covered in ice, to better understand the relationship between CO2 and global temperature. It finds that while most future warming ...
East coast mussel shells are becoming more porous in warming waters
2024-04-17
Researchers at the American Museum of Natural History have found that over the last 120 years, the porosity—or small-scale holes—in mussel shells along the East Coast of the United States has increased, potentially due to warming waters. The study, which analyzed modern mussel shells in comparison to specimens in the Museum’s historic collection, was published today in the journal PLOS ONE.
“Mussels are important on so many levels: they provide habitats on reefs, they filter water, they protect coasts during storms, and they are important commercially as well—I love mussels and I know many other people do, too,” said Leanne Melbourne, ...
Paleontologists unearth what may be the largest known marine reptile
2024-04-17
The fossilised remains of a second gigantic jawbone measuring more than two metres long has been found on a beach in Somerset, UK.
Experts have identified the bones as belonging to the jaws of a new species of enormous ichthyosaur, a type of prehistoric marine reptile. Estimates suggest the oceanic titan would have been more than 25 metres long.
Father and daughter, Justin and Ruby Reynolds from Braunton, Devon, found the first pieces of the second jawbone to be found in May 2020, while searching for fossils on the ...
Video-assisted hand therapy is effective after thumb arthritis surgery
2024-04-17
Waltham — March 15, 2024 — For patients undergoing carpometacarpal (CMC) joint surgery for treatment of thumb osteoarthritis, the use of online video instruction for postoperative hand therapy is associated with outcomes similar to in-person therapy visits – while substantially reducing travel time and distance, reports a clinical trial in The Journal of Bone & Joint Surgery. The journal is published in the Lippincott portfolio in partnership with Wolters Kluwer.
Video-administered hand therapy provides an "efficient and effective" alternative to in-person therapy after CMC arthroplasty, ...
New butterfly species created 200,000 years ago by two species interbreeding
2024-04-17
Researchers have shown that an Amazonian butterfly is a hybrid species, formed by two other species breeding together almost 200,000 years ago.
The discovery, by an international team led by scientists at the University of York and Harvard University, demonstrates how the formation of new species can be more complex than previously imagined.
Species are often thought of as the tips, or leaves in a ‘tree of life’. In this model, new species are produced by the tips splitting over thousands to millions of years.
Scientists now understand, however, that the branches in the tree of life are ...
New class of antimicrobials discovered in soil bacteria
2024-04-17
Researchers have discovered toxic protein particles, shaped like umbrellas, that soil bacteria known as Streptomyces secrete to squelch competitors, especially others of their own species.
The discovery of the umbrella toxin particles and related information about their structures, composition and mode of action were published April 17 in Nature.
The umbrella toxin proteins are the latest example of these bacteria’s varied, combative strikes on their microscopic rivals. The crowded, diverse bacteria communities ...
Substantial global cost of climate inaction
2024-04-17
Traditionally, estimates of how climate change will affect global economies have focused on the effects of annual temperature changes. However, the additional impacts of variability and extremes in rainfall and temperature have remained largely unexplored, until now. Using projections from 33 global climate models, an international research team, led by Paul Waidelich at ETH Zurich, conducted a pioneering study, published in the journal, Nature Climate Change, to quantify such impacts on gross domestic product (GDP) across the globe.
Revealing the additional ...
Mountain chickadees have remarkable memories. A new study explains why
2024-04-17
Lost your keys? Can’t remember where you parked the car? If only you had the memory of a mountain chickadee.
These half-ounce birds, with brains slightly larger than a pea, stash tens of thousands of food items like seeds in tree bark, under dead leaves and inside pinecones across the mountains. When winter arrives, they can recall the exact locations of their caches, a skill that helps them survive the bitter cold and deep snow.
In a new study published April 17 in the journal Current Biology, researchers at the University of Colorado Boulder and the ...
Current police response to intimate partner violence calls for change
2024-04-17
April 17, 2024-- Policing of intimate partner violence (IPV) may result in adverse consequences for survivors, according to a new study at Columbia University Mailman School of Public Health. However, the evidence concerning the generalized consequences of IPV policing has not been comprehensively evaluated until now and the results call into question whether IPV policing benefits survivors.
This is the first review on the consequences of IPV policing in the U.S. The findings are published in the journal Aggression and Violent Behavior.
IPV, which includes physical violence, sexual violence, psychological abuse, and other forms of coercion between current or former spouses or ...
Understanding climate warming impacts on carbon release from the tundra
2024-04-17
The warming climate shifts the dynamics of tundra environments and makes them release trapped carbon, according to a new study published in Nature. These changes could transform tundras from carbon sinks into a carbon source, exacerbating the effects of climate change.
A team of over 70 scientists from different countries used so called open-top chambers (OTCs) to experimentally simulate the effects of warming on 28 tundra sites around the world. OTCs basically serve as mini-greenhouses, blocking wind and trapping heat to create local warming.
The warming experiments led to a 1.4 degrees Celsius increase in air temperature and a 0.4 degrees increase ...