Realising the real-world benefits of AI in healthcare

Published August 2025 | 5 min read
Expert contributors: Professor Farah Magrabi, Co-Chair of the Safety, Quality and Ethics Working Group, Australian Alliance for Artificial Intelligence in Healthcare, and Professor of Biomedical and Health Informatics at the Australian Institute of Health Innovation, Macquarie University; Dr Chris Pettigrew, Head of the HCF Research Foundation
Words by Beth Wallace

The use of AI in healthcare could transform how we prevent, diagnose and treat illness. Here’s the latest research making that happen.

Artificial intelligence (AI) is quickly becoming a normal part of everyday life. Broadly defined as “computer systems able to perform tasks normally requiring human intelligence”, AI is regularly used by 50% of Australians, and is currently being adopted by 40% of small- and medium-size businesses. From virtual assistants and chatbots to image generators and writing tools, AI is now being used across a wide range of industries and professions.

AI has the potential to transform healthcare in many ways. Beyond speeding up admin tasks, it can also be used to improve diagnostic accuracy and personalise treatment.

Already, GPs have been quick to adopt tools like AI scribe for note-taking tasks and patient health summaries, but AI technologies have much more to offer, says research leader in AI systems Professor Farah Magrabi.

Prof Magrabi is Co-Chair of the Safety, Quality and Ethics Working Group at the Australian Alliance for Artificial Intelligence in Healthcare, and Professor of Biomedical and Health Informatics at the Australian Institute of Health Innovation, Macquarie University.

“There’s a lot of excitement about how AI can improve healthcare, and there are many efforts underway to make sure it’s used safely and responsibly,” she says. “We feel like it's still the early days and the best is yet to come.”

How is AI used in healthcare?

There are many ways AI can be used in healthcare settings. These include:

• image analysis (e.g. cancer screenings, skin checks)
• genetic data analysis (e.g. to diagnose and treat genetic disease)
• robots for companionship and remote monitoring (e.g. in aged care)
• clinical decision support
• chatbots
• surgical tools
• medical record analysis (e.g. to identify risk, personalise treatment).

Prof Magrabi says these and other AI applications could benefit the healthcare sector in a number of ways, including by reducing the burden on healthcare resources, making better use of data and improving the patient experience.

One aspect of healthcare that’s already ticking these boxes is medical imaging.

For radiologists, oncologists, gastroenterologists and other medical professionals who rely on images for diagnosis, AI can serve as a ‘smart assistant’, analysing images and spotting details human eyes can miss. It can be trained to do this by studying thousands of images; the more these AI assistants ‘see’, the better they become at spotting patterns, like the early signs of cancer.

“You need lots of good-quality data to develop these systems,” Prof Magrabi explains. “And medical images have always been standardised, which has given radiology [and other departments that use medical imaging] a head start.”

What does this mean for the Australian healthcare system?

With ongoing funding and research, Prof Magrabi believes Australia is in a strong position to expand the use of AI in healthcare, with two factors setting us apart from many other countries.

“One is that we have undertaken a citizens’ jury” [a group of randomly selected people representative of the wider population who make recommendations on an issue]: a democratic process to shape how AI should be used in healthcare,” she says.

“The second is we have A National Policy Roadmap for Artificial Intelligence in Healthcare, which sets out what we want to do and how we can do it safely and effectively.”

Prof Magrabi says this roadmap – and soon-to-be-released guidelines from the Australian Commission on Safety and Quality in Health Care – aim to address some of the challenges linked to the safe and responsible use of AI in healthcare. These include:

• Safe use: Clinicians must familiarise themselves with the intended use of each AI tool, understanding the benefits and potential harms. AI tools intended to be used for medical purposes like diagnosis and treatment are regulated.
• Accountability: Always review AI-generated outputs for accuracy and quality. Prof Magrabi says human checking is crucial, and clinicians are ultimately responsible for decision-making.
• Privacy: Steps need to be taken to protect patient data that’s used to train AI systems.
• Equity: AI systems can develop biases if they’re trained on datasets that don’t reflect real-world populations. This can have a negative impact on disadvantaged or marginalised groups.

AI research funded by the HCF Research Foundation

With so many potential uses and benefits of AI in healthcare, the HCF Research Foundation is investing in the growing body of research behind it.

Since launching in 2000, the Foundation has funded health and medical research for the benefit of members and all Australians. One of its initiatives, the Translational Research Grant (TRG) scheme, supports the adoption of cost-effective, evidence-based innovations in healthcare to improve patient outcomes and enhance the work experience of clinical staff.

Translational research involves taking findings from clinical studies and introducing them to real-world community or health settings, explains Dr Chris Pettigrew, Head of the HCF Research Foundation. “The Translational Research Grants are really about taking research that works, and evidence we already have, and implementing it into practice.”

For each TRG funding round, the Foundation selects a different focus area, supporting projects that target specific aspects of healthcare. The 2024 TRG round called for applications addressing the topic of “embedding AI into diagnostic and clinical care pathways”.

“What we've observed is that the use of AI in healthcare has been largely administrative, whereas the use of AI in clinical decision-making and in care pathways is a really important application, and that's what we wanted to focus on with this grant round,” says Dr Pettigrew. “we were interested in how AI can help the actual care delivery.”

Dr Pettigrew explains how four TRG recipients are using their funding.

1. Preventing blindness in pregnant women with diabetes

Diabetic retinopathy is a serious eye condition that can lead to blindness. It affects one in three people with diabetes (types 1 and 2) – and the risk for women with pre-gestational diabetes (pre-existing diabetes) increases with pregnancy.

To help people at risk get the specialist care they need, the Centre for Eye Research Australia is trialling AI-assisted cameras in obstetric clinics. While patients are waiting to see their obstetrician, or immediately after their appointment, they can get their eyes checked.

Dr Pettigrew says this is an important project because it helps pregnant women access care to support the early detection of disease. “Pregnant women with diabetes have a lot of appointments to manage, and that's on top of their regular life commitments and everything else going on,” he says. “This trial brings the diagnostics to them in a place where they already need to be, and they've got the time to do it.”

2. Improving access to cardiac scans

Aortic stenosis is a serious (and potentially fatal) heart condition that affects your aortic heart valve. An echocardiogram, or cardiac ultrasound, diagnoses the condition, but access to this equipment – and the experts who operate it and interpret the results – can be limited, particularly in rural and remote parts of Australia.

A team from the University of Tasmania is trying to overcome both challenges with the help of AI.

“AI can help GPs and registered nurses, people who aren't specifically trained as ultrasound technicians, to use portable ultrasound machines to take images of the heart, guiding them to get the best images,” says Dr Pettigrew. “Then, the images get sent off and are read by a different AI, which helps the clinician make a diagnosis. So, it's AI in the acquisition and in the interpretation of those images.

“If successful, it will mean patients living in rural and regional settings don't need to travel to a large facility – they can just go to their GP.”

3. Speeding up the diagnosis of cardiovascular disease

A team from Northern Health in Melbourne’s outer north is working to make diagnosing cardiovascular disease (CVD) more efficient by introducing a cardiac point-of-care ultrasound (POCUS) to hospital emergency departments.

The AI-powered tool performs a “quick ultrasound” and can be used by non-specialists (e.g. emergency department staff) on patients experiencing chest pain, explains Dr Pettigrew. “It can lead to a much quicker diagnosis, with the AI helping clinicians in the ED to capture and then interpret the ultrasound images, rather than having to wait for a slot in the imaging department and a specialist to read the images.”

This should improve patient flow in emergency departments, he adds, making sure people “get the care they need when they need it” – reducing complications and improving survival rates.

4. Overcoming language barriers in hearing tests

Hearing tests in Australia are typically delivered in English, which doesn't serve our culturally and linguistically diverse population. To address this, the National Acoustic Laboratories have developed ULI+ – an AI-powered speech test that delivers word sequences audio-visually to assess hearing levels.

“It’s language-neutral,” says Dr Pettigrew. “They’ve tested it across more than 40 languages and the results are better than an English-based test.”

This trial will establish whether offering this speech test in typical hearing centres around the country helps people get equitable, effective and personalised hearing care.

Dr Pettigrew adds, “It's about helping people with hearing loss, regardless of their language background.”

Reshaping the future of healthcare

For 25 years, the HCF Research Foundation has supported innovative, translational research that aims to improve healthcare outcomes for all Australians. We do this by funding research projects that focus on new ways to deliver health services and provide high-quality, affordable care when and where it is needed.

Read more about the history and vision of the HCF Research Foundation.