What does AI mean for healthcare in Australia?

What does AI mean for healthcare in Australia?

Subscribe to the L&D Toolbox

Artificial intelligence is hardly a new frontier: Australia has already set itself up as a leader in this innovative new area of technological solutions, having revolutionised the way healthcare workers and professionals detect cancer and COVID-19, or treat anxiety disorders (AGDISER, 2021).

However, given the expected boom – a functional AI industry could bring around $315B to our economy by 2028 (AGDISER, 2021) – there are also expected to be a flurry of new innovations that will change the way the Australian healthcare system functions. These expectations of growth take into account the fact that the innovation is not the whole project: there must be someone on the other end who is able to use it.

So what’s the plan to both foster AI innovation in health, and also improve digital literacy to ensure its ease of use?

First, let’s define the scope of what actually constitutes artificial intelligence.

What is artificial intelligence?

According to the Australian Government Department of Industry, Science, Energy and Resources (AGDISER, 2021):

AI is a collection of interrelated technologies that can be used to solve problems autonomously and perform tasks to achieve defined objectives. AI is more than just the mathematical algorithms that enable a computer to learn from text, images or sounds. It is the ability for a computational system to sense its environment, learn, predict and take independent action to control virtual or physical infrastructure.

If you’re still a bit hazy on the actual, real-world applications of the above definition, the following is a list of examples of artificial intelligence that you probably encounter on a daily basis:

  • Online maps and navigation (especially when optimised for smartphones)
  • Facial recognition to unlock phones
  • Search and recommendation algorithms - this includes Ausmed’s sophisticated recommendations system, which sees what you’re learning and gives you more resources you may enjoy
  • Siri or Alexa
  • Social media content and ‘for you’ page content.

The long and short of it is that artificial intelligence, from a very macro point of view, is trying to mimic human cognitive function but without the fault of ‘human errors’ (Jiang et al, 2017).

How could this technology revolutionise healthcare?

As has already been said, AI has already made an impact on data interpretation to help detect COVID-19. In terms of how this was done, AI algorithms minimised the spread of the virus in Australia by identifying hotspots and tracing relevant contacts and prompting workers to tell certain people to isolate (Scott et al, 2020). This was a great way to free up public health resources, and was an easily-applicable system that was reactionary and timely (Scott et al, 2020).

“As the COVID-19 pandemic has taught us all, business as usual in the 21st Century is business as change. From research bench to clinical bedside and into the hands of patients, AI promises to make Australian healthcare a learning system that is more nimble, adaptive, personalised, safe and effective.” - Professor Enrico Coiera, leader of the Australian Alliance for Artificial Intelligence in Healthcare (Waldie, 2021)

However, how does AI help detect cancer? Essentially, very complex AI algorithms are able to reduce the frequency of false positive detections and can classify unusual tumours using ‘methylation analysis’ (Savage, 2020). The methylation method requires a full genome-wide methylation profile (it checks for small hydrocarbons attached to the DNA strand), which is then fed into the AI algorithm and tested against other methylation profiles in the system. This allows the system to align the current profile with the most likely type of tumour present, and also provides a ‘confidence score’ (Savage, 2020).

Exciting AI projects just past or on the horizon

The economic boom that pharmaceutical companies enjoyed during the COVID-19 pandemic brought about a growth in AI-related deals and purchases globally (Lo, 2021). More specifically, the value of AI-related deals in Q3 2019 was around $2,805M, while in Q3 2020 the value had soared to $21,320M (Lo, 2021). That’s almost seven times more value accrued in a single year. As a result, the next few years should see a marked growth in high-quality AI-related technology being produced worldwide.

While it’s unclear how many of the deals that made up that $21,320M were set in Australia, there are some exciting research projects occurring already. A collection of Australian organisations and universities are working together to develop an AI system that can scan for and detect brain aneurysms faster and more accurately (AGDISER, 2021). The goal of this technology is to help radiologists with early detection and maintain a high quality of life for as long as possible for people who experience aneurysms. While the AI technology will be crafted to suit this specific project, the system can be generalised and will push AI systems development and innovation in Australia.

How is Australia going to create a culture of AI in health?

Digital literacy

For Australia to confidently move towards a strong culture of AI in healthcare, we must support older Australians to stay involved. Approximately 34% of Australians aged over 50 years are either digitally disengaged or have low digital literacy levels (eSafety Commissioner, 2018). Thankfully, there are many organisations whose missions are to help older Australians stay engaged with technology and, by definition, their health: Be Connected, eSafety.gov, the Australian Digital Inclusion Index are all examples of this.

However, it’s not only older Australians who may struggle with technological ability; many people of all ages and abilities lack intuitive digital skills. The National Digital Health Strategy aims to make new digital resources and systems as easy-to-use and accessible for both patients and those who work within healthcare organisations (Digital Health, 2021).

One of the biggest stressors when someone needs medical attention is that they don’t fully understand what’s happening, and how their care team is ensuring best outcomes. This unsureness is largely due to confusing technological advancements that older adults have never used and don’t understand. As the government’s digital literacy campaigns slowly increase confidence regarding technology, older Australians will be in a much better position to understand the complex technology that is being used by their care team.

In the end, Australia’s burgeoning AI technology is only as useful as it is usable for workers and understandable for patients.

Funding

Funding is one of the biggest blockers to AI development: not only do skilled AI innovators have to be paid a competitive salary to keep them employed in Australia, the actual implementation of new systems can be expensive. Regarding the cancer-screening methylation method mentioned previously, that technology is so expensive only a small group of hospitals in the US can afford the methylation profiling equipment that the AI algorithm relies upon (Savage, 2020). That being said, Australia’s Medical Research Future Fund (MRFF) receives regular grants from the federal government that are put directly towards research projects furthering AI (AGDISER, 2021).

The fight doesn’t end with funding, however: safety testing and longitudinal studies mean that the journey from ideation to implementation is a long and arduous one, especially for medical technologies that utilise something as new and unprecedented as AI (Kuan, 2019).

Governance

The Australian Government empowers businesses to create and assign their own governance frameworks within their AI development schemes. However, the Australian Human Rights Commission’s ‘Artificial Intelligence: governance and leadership white-paper (2019)’ provides a loose guide on how to create and implement governance standards (Davis, 2019).

Additionally, the AI Ethics Framework that was founded by the government in 2019 is a great display of commitment to standards and safety regarding AI development in Australia. This framework contains 8 principles, each of which aim to reduce the negative impacts of AI – such as job displacement and exclusivity based on gender, race or ability – and ensures governance is implemented and abided by (AGDISER, 2021).

However, any guides on how to implement and monitor AI frameworks in Australia are mostly based upon assumptions and past performance in other countries (though that past performance is still relatively recent). For example, the EU is working hard to stay in the AI race with China and the US by creating a watertight and streamlined governance system that optimises the region’s rate of production (Stix, 2021).

What now?

You’ve now got a basic understanding of how AI will affect Australia’s healthcare sector – and how it’s already helping some people receive higher-quality care. But what can you do next?

A great way to stay updated on the Australian AI sector, and current healthcare projects, is to read through the Department of Industry, Science, Energy and Resources’ ‘Implementation and Next Steps’ section of the Australia’s AI Action Plan 2021 document (AGDISER, 2021). This sets out the plan for budgets, governance, ideal projects and industry revenue.

You can also follow the Australian Alliance for Artificial Intelligence in Healthcare on Twitter. The AAAiH are a guiding force in Australia’s AI development, and are a great source of reliable and breaking news in the industry.

Additionally, you can use Ausmed’s educational resources exploring the current impacts of virtual reality technology (VR) in healthcare. While not exactly artificial intelligence, VR is a neighbouring technology and both areas will work together to provide sophisticated solutions to complex health problems in the future:


References