5 ways AI is transforming healthcare

This article has been updated.

With 4.5 billion people currently without access to essential healthcare services and a health worker shortage of 10 million expected by 2030, AI has the potential to help bridge that gap and revolutionize global healthcare.

It could even get us back on track to meet the United Nations' Sustainable Development Goal of achieving universal health coverage by 2030.

But while the technology is rapidly developing, healthcare is "below average" in its adoption of AI compared to other industries, according to the World Economic Forum's latest white paper, The Future of AI-Enabled Health: Leading the Way.

"AI transformation goes beyond adopting new tools," it says. "It involves rethinking the fundamentals of how health is delivered and accessed."

Here are five ways AI is already transforming healthcare.

A new AI software is "twice as accurate" as professionals at examining the brain scans of stroke patients. Two UK universities trained the software on a dataset of 800 brain scans of stroke patients and then trialled it on 2,000 patients.

The results were impressive. Alongside the AI model's accuracy, the software was also able to identify the timescale within which the stroke happened - crucial information for professionals.

As Dr Paul Bentley, consultant neurologist, told the Health Tech Newspaper: “For the majority of strokes caused by a blood clot, if a patient is within 4.5 hours of the stroke happening, he or she is eligible for both medical and surgical treatments. Up to 6 hours, the patient is also eligible for surgical treatment, but after this time point, deciding whether these treatments might be beneficial becomes tricky, as more cases become irreversible. So it’s essential for doctors to know both the initial onset time, as well as whether a stroke could be reversed.”

Surprisingly, urgent care doctors miss broken bones in up to 10% of cases. What’s more, X-ray technicians are both in short supply and overloaded.

So using AI to do the initial scan could potentially avoid both unnecessary X-rays and missed fractures. The UK’s National Institute for Health and Care Excellence (NICE) says the technology is safe, reliable and could reduce the need for follow-up appointments.

But there are concerns around the fast rollout of AI in healthcare.

"It is important that people using these tools are properly trained in doing so, meaning they understand and know how to mitigate risks from technological limitations ... such as the possibility for wrong information being given," Dr Caroline Green of the Institute for Ethics in AI at the University of Oxford told the BBC.

In the UK, around 350,000 people are taken by ambulance to hospital each month. It’s down to paramedics to decide who does or doesn’t need to go, and always with an awareness of how few beds are available.

A study in Yorkshire in the north of England found that in 80% of cases AI could correctly predict the patients that needed to be transferred to hospital. The AI model was trained on factors such as a patient’s mobility, pulse and blood oxygen levels and chest pain – it also proved to respond without bias. NICE did caution, however, that before being put into more widespread use, more training was needed.

A new AI machine learning model can detect the presence of certain diseases before the patient is even aware of any symptoms, according to its maker AstraZeneca.

Using medical data from 500,000 people who are part of a UK health data repository, the machine could “predict with high confidence a disease diagnosis many years later”.

Slavé Petrovski, who led the research, told Sky News: "For many of these diseases, by the time they manifest clinically and the individual goes to the doctor because of an ailment or visible observation, that is far down the line from when the disease process began.

“We can pick up signatures in an individual that are highly predictive of developing diseases like Alzheimer's, chronic obstructive pulmonary disease, kidney disease and many others," he said.

Doctors must make informed swift medical decisions: AI could potentially speed these up, but it could also provide unreliable or biased information.

A US study found that standard large language models (LLMs) like ChatGPT, Claude or Gemini were unable to provide clinicians with sufficiently relevant or evidence-based answers to their medical questions. But ChatRWD, a retrieval-augmented generation (RAG) system – which essentially combines LLMs with retrieval systems to improve output – produced useful answers to 58% of the questions (compared with 2%-10% for the LLMs).

Digital interfaces are increasingly being rolled out to help triage patients, too. In an insight report from 2024, part of the World Economic Forum’s Digital Healthcare Transformation Initiative, a case study on digital patient platform Huma, revealed it could reduce readmission rates by 30%, time spent reviewing patients by up to 40% and “alleviated the workload of healthcare providers”.

The report anticipates a future in which technologies like these could “dramatically transform the patient experience. People who are generally healthy can use self-monitoring devices to optimize their mental and physical health, while those with health issues will have access to a wide range of digital solutions”.

To ensure this is done equitably and safely, regulation of AI tools is key. In the UK, AI-powered medical devices are strictly regulated by the Medicines and Healthcare products Regulatory Agency. In the US, the Food and Drug Administration (FDA) last year examined the regulation of AI in healthcare and concluded that, while the FDA will “continue to play a central role in ensuring safe, effective, and trustworthy AI tools” it was also essential that “all involved entities … attend to AI with the rigour this transformative technology merits”.