(Press-News.org) A new study in the academic journal Machine Learning: Health discovers that ChatGPT can accelerate patient screening for clinical trials, showing promise in reducing delays and improving trial success rates.
Researchers at UT Southwestern Medical Centre used ChatGPT to assess whether patients were eligible to take part in clinical trials and were able to identify suitable candidates within minutes.
Clinical trials, which test new medications and procedures on the public, are vital for developing and validating new treatments. But many trials struggle to enrol enough participants. According to a recent study, up to 20% of National Cancer Institute (NCI)-affiliated trials fail due to low enrolment. This not only inflates costs and delays results, but also undermines the reliability of new treatments.
Currently, screening patients for trials is a manual process. Researchers must review each patient’s medical records to determine if they meet eligibility criteria, which takes around 40 minutes per patient. With limited staff and resources, this process is often too slow to keep up with demand.
Part of the problem is that valuable patient information contained in electronic health records (EHRs) is often buried in unstructured text, such as doctors’ notes, which traditional machine learning software struggles to decipher. As a result, many eligible patients are overlooked because there simply isn’t enough capacity to review every case. This contributes to low enrolment rates, trial delays and even cancellations, ultimately slowing down access to new therapies.
To counter this problem, the researchers have looked at ways of speeding up the screening process by using ChatGPT. Researchers used GPT-3.5 and GPT-4 to analyse 74 patients’ data to see if they qualified for a head and neck cancer trial.
Three ways of prompting the AI were tested:
Structured Output (SO): asking for answers in a set format.
Chain of Thought (CoT): asking the model to explain its reasoning.
Self-Discover (SD): letting the model figure out what to look for.
The results were promising. GPT-4 was more accurate than GPT-3.5, though slightly slower and more expensive. Screening times ranged from 1.4 to 12.4 minutes per patient, with costs between $0.02 and $0.27.
“LLMs like GPT-4 can help screen patients for clinical trials, especially when using flexible criteria,” said Dr. Mike Dohopolski, lead author of the study. “They’re not perfect, especially when all rules must be met, but they can save time and support human reviewers.”
This research highlights the potential for AI to support faster, more efficient clinical trials - bringing new treatments to patients sooner.
The study is one of the first articles published in IOP Publishing's Machine Learning series™, the world’s first open access journal series dedicated to the application and development of machine learning (ML) and artificial intelligence (AI) for the sciences.
The same research team have worked on a method that allows surgeons to adjust patients’ radiation therapy in real time whilst they are still on the table. Using a deep learning system called GeoDL, the AI delivers precise 3D dose estimates from CT scans and treatment data in just 35 milliseconds. This could make adaptive radiotherapy faster and more efficient in real clinical settings.
ENDS
About IOP Publishing
IOP Publishing is a society-owned scientific publisher, delivering impact, recognition and value to the scientific community. Its purpose is to expand the world of physics, offering a portfolio of journals, ebooks, conference proceedings and science news resources globally. IOPP is a member of Purpose-Led Publishing, a coalition of society publishers who pledge to put purpose above profit.
As a wholly owned subsidiary of the Institute of Physics, a not-for-profit society, IOP Publishing supports the Institute’s work to inspire people to develop their knowledge, understanding and enjoyment of physics. Visit ioppublishing.org to learn more.
About the lead author
Dr. Dohopolski is a physician-scientist in radiation oncology at UT Southwestern with experience integrating artificial intelligence (AI) into clinical practice, notably developing advanced predictive models and pioneering the use of large language models (LLMs) for clinical trial screening. He received an engineering degree from Notre Dame and a medical degree from the University of Pittsburgh School of Medicine. During residency at UT Southwestern Medical Center, he completed the competitive Holman Pathway, dedicating 21 months exclusively to AI-focused research under mentors Drs. Jing Wang and Steve Jiang from the Medical Artificial Intelligence and Automation (MAIA) lab. This currently work has now been implemented in a prospective manner across several studies across the university.
END
Are you eligible for a clinical trial? ChatGPT can find out
2025-07-31
ELSE PRESS RELEASES FROM THIS DATE:
New treatment could reduce brain damage from stroke, study in mice shows
2025-07-31
Cambridge scientists have developed and tested a new drug in mice that has the potential to reduce damage to the brain when blood flow is restored following a stroke.
As many as one in four people will have a stroke during their lifetime. This is when a blood clot prevents oxygen from reaching a part of the brain. The first few hours following a stroke are crucial – the blood clot needs to be removed quickly so that the oxygen supply to the brain can be restored; otherwise, the brain tissue begins to die.
Currently, the outcome for stroke patients receiving even the best available ...
4,000-year-old teeth record the earliest traces of people chewing psychoactive betel nuts
2025-07-31
In south-east Asia, betel nut chewing has been practiced since antiquity. The plants contain compounds that enhance the consumer’s alertness, energy, euphoria, and relaxation. Although the practice is becoming less common in modern times, it has been deeply embedded in social and cultural traditions for thousands of years. Chewing betel nuts typically results in dark, reddish-brown to black stained teeth.
Yet, teeth without staining may not mean that people didn’t chew betel nuts. Now, using a new method, an international team ...
Efficient solar harvesting even in high humidity
2025-07-31
The Energy & Environment Materials Research Division of the Korea Institute of Materials Science (KIMS), led by Dr. Dong-chan Lim and Dr. So-yeon Kim, has developed a highly durable flexible perovskite solar cell material and fabrication process that remains stable even under high humidity conditions. This breakthrough enables the production of high-efficiency solar cells in ambient air without the need for expensive equipment, offering the potential for significant cost reductions in manufacturing.
Perovskite has attracted attention as a next-generation material capable of replacing conventional silicon solar cells due to its excellent light absorption, ...
Heavy drinking raises the risk of undesired pregnancy; cannabis use does not
2025-07-31
A new study has found that, among women with a high desire to avoid becoming pregnant, those who drank heavily had a 50% higher risk of becoming pregnant than those who drank moderately or not at all. In contrast, participants who used cannabis were no more likely to have an undesired pregnancy than participants who did not use cannabis.
From a larger sample of over 2,000 non-pregnant women aged 15-34, researchers identified a subgroup of 936 who didn't want to get pregnant. Within that subgroup, ...
New study shows young adults who use high strength cannabis do not ‘titrate’ to less risky levels of use
2025-07-31
A new study published today in the scientific journal Addiction has found that young adults in the US do not ‘titrate’ when using strong cannabis. In other words, they do not use less cannabis to compensate for the stronger potency. In fact, it’s the opposite: young adults who report using strong cannabis also typically use it more frequently and in higher quantities than young adults who use weaker forms of the drug.
This study surveyed over 400 young adults in California USA who had used cannabis in the past month. They were asked to describe the strength of their typical cannabis product using a THC potency scale of 0 (no THC) to 5 (very high THC concentration). ...
Black hole vibes
2025-07-31
Kyoto, Japan -- Black holes embody the ultimate abyss. They are the most powerful sources of gravity in the universe, capable of dramatically distorting space and time around them. When disturbed, they begin to "ring" in a distinctive pattern known as quasinormal modes: ripples in space-time that produce detectable gravitational waves.
In events like black hole mergers, these waves can be strong enough to detect from Earth, offering a unique opportunity to measure a black hole's mass and shape. However, precise calculation of these vibrations through theoretical methods ...
Actual distance travelled by migrating whales drastically underestimated
2025-07-31
Whales are swimming further than previously estimated - up to 20% more - according to a new study published in Ecology.
Co-authored by Griffith University’s Dr Olaf Meynecke, the study reveals traditional methods for calculating animal movement may be drastically underestimating the actual distance travelled, particularly for marine species such as whales.
“For years, we've tracked whales using satellite tags, plotting their movements across oceans,” Dr Meynecke said.
“But this research shows we’ve been looking ...
The eagles resistant to poisonous toads
2025-07-31
Kyoto, Japan -- In 1978, cane toads, which are native to South and Central America, were introduced for pest control to Ishigaki island in Okinawa prefecture in Japan. These poisonous toads secrete deadly toxins, killing enough of the predators in their new territories that they have been designated an invasive species.
However, one predator on Ishigaki has proven resistant to the cane toads' poison. The crested serpent eagle, common across Asia but considered critically endangered in Japan where only about 200 individuals reside, has been observed feeding on the toads on the ...
Cyberstalking growing at faster rate than other forms of stalking
2025-07-31
Cyberstalking is increasing at a faster rate than traditional stalking and is disproportionately affecting young people, women, and members of the lesbian, gay and bisexual community, according to a new study led by researchers from UCL.
The study, published in the British Journal of Criminology, is the first to use nationally representative data from the Crime Survey for England and Wales (CSEW) to examine the prevalence and perception of cyberstalking over an eight-year period (2012–2020).
It revealed that while cyberstalking remains less common than physical stalking, the proportion of ...
CPADS: a web tool for comprehensive pancancer analysis of drug sensitivity
2025-07-30
CPADS integrates data from the Gene Expression Omnibus (GEO), The Cancer Genome Atlas (TCGA), and the Genomics of Drug Sensitivity in Cancer (GDSC) databases, encompassing over 29,000 samples across 44 cancer types and involving 288 drugs. It provides five main analysis modules: differential expression analysis, correlation analysis, pathway analysis, drug analysis, and gene perturbation analysis. These modules enable users to explore gene expression changes, correlations between genes or drugs, pathway enrichment, drug sensitivity, and the ...