Key points
- According to a new report by our e-Health Research Centre, the use of artificial intelligence (AI) in healthcare will continue to increase.
- The adoption of AI in areas such as clinical decision support and administrative tasks will become more common.
- Our scientists are using AI to facilitate Alzheimer’s disease research, reduce clinician burnout, and enhance genomic data sharing and analysis.
There are many areas of daily living that utilise artificial intelligence (AI) systems for efficiency, improved performance, and convenience. However, when it comes to healthcare the accuracy and safety of AI can sometimes mean the difference between life and death. An important new report by our Australian e-Health Research Centre (AEHRC) explores the challenges and the opportunities posed by the adoption of AI in healthcare.
The report, titled AI Trends for Healthcare, shows how AI systems may be able to provide patients with increased ownership and control over their data. Simultaneously, it could provide safe and secure data analytics to help governments across the world prevent and control pandemics.
The report is jam packed with other examples of how AI may affect your health. So, we thought we’d give you a taste of what’s there by starting with the basics. Which means the ABCs, of course.
A is for Alzheimer’s disease research
Dementia is currently the second leading cause of death for all Australians. It is soon predicted to be the leading cause of death.
Alzheimer’s disease (AD) is the most common form of dementia. Research institutions across the globe are conducting research on preventing and treating AD.
Large studies of patients with AD have produced massive amounts of data. These studies often involve different cognitive tests and scales. This is because analysing and drawing conclusions from the data can be challenging.
Rosita Shishegar and her colleagues at the AEHRC came up with a solution. They developed an AI-based method for harmonising data across AD studies. This has enabled the creation of the largest dataset of Alzheimer’s disease in the world.
By bringing the data into alignment, scientists can use the information to make more accurate predictions.
Since the initial results were published, the team have extended the method to other cohorts. They are part of an international, multi-centre project that aims to track cognitive decline over time and identify individuals at risk of developing AD.
AI has more to offer Alzheimer’s research. Rosita explained that having access to large datasets means we can use AI methods to more accurately evaluate AD data.
"AI is the same as any human in how it learns. It can only make predictions based on the data it’s given,” Rosita said.
“The more data we have on AD, the more we can leverage AI technologies to evaluate risk factors and make predictions.”
B is for reducing clinician burnout
AI software and humans have different strengths. AI tools excel at processing large amounts of data and accurately completing repetitive and monotonous tasks. Humans are good at seeing things holistically and can connect with one another in a way that AI can’t replicate.
Clinicians spend a lot of time on day-to-day administrative tasks like documentation. These tasks, while important, add to clinicians’ already high workload and contribute to burnout.
AI is perfectly suited to these kinds of tasks. AI methods can allow appropriate data to be provided to clinicians at the point of care. Clinicians can spend less time sifting through information and more time working directly with patients.
An area of healthcare set to benefit from AI is medical imaging. AEHRC researchers are using machine learning, a type of artificial intelligence, to develop new medical imaging tools.
Medical images are interpretated by trained experts. Even so, interpreting it is time-consuming and complicated. Our researchers are creating AI software that can analyse and make predictions from medical images. These tools could potentially reduce the burden on clinicians while helping them diagnose diseases earlier and diagnose conditions more accurately.
C is for cloud computing
Cloud computing uses remote servers hosted on the internet to store, manage, and process data. Health systems are increasingly using the technology to store and exchange data securely and efficiently.
AEHRC researchers are combining the benefits of cloud computing with AI technology. Dr Denis Bauer leads the Transformational Bioinformatics group. They are world leaders in cloud-native genomics research.
They developed VariantSpark, a tool that utilises machine learning to detect changes in the SARS-CoV-2 genome. SARS-CoV-2 is the virus that causes COVID-19.
Researchers were able to analyse the whole genomes of 10,520 SARS-CoV-2 samples. This was more than had ever been analysed previously. It meant they could better detect dangerous and emerging COVID-19 variants.
Denis' group also developed sBeacon. This data-sharing system enables researchers from around the world to share genomic information.
The technology is part of an emerging area of healthcare called precision medicine. It involves analysing an individual’s genomic data and comparing it to population-level genomic data. It can help diagnose diseases and improve treatment.
Denis said that they developed sBeacon to facilitate this process.
“sBeacon can rapidly find and isolate the disease-causing genomic mutations amongst the three billion letters of the human genome,” she said.
“With this technology, we can potentially improve the treatment of complex diseases like cancer and cardiovascular disease.”
Because the technology is based in the cloud, it's around 10 times cheaper and over 1,000 times faster than existing software.
These aren’t the only uses for AI in healthcare. From Alzheimer’s disease research to zeroing in on outbreaks of antimicrobial resistant pathogens, there’s a whole alphabet of AI-driven innovation waiting to be explored.