FEATURE ARTICLES
AI IS MAKING THE IMPOSSIBLE POSSIBLE IN HEALTHCARE By Robert Merlicek Chief Technology Officer of TIBCO – APJ
Artificial intelligence (AI) has the potential to completely reshape Australia’s healthcare system by enabling more efficient delivery of higher-quality services. AI enables data analytics: the gathering of insights from data that would be otherwise impossible for a human to assimilate and analyse. Data analytics enables sense to be made of the large datasets generated by healthcare organisations. Pfizer, for example, is transforming its manufacturing processes to be more predictive and adaptive. Pfizer’s Director of Manufacturing Intelligence, Hamid Mehdizadeh, says: “While our scientists push the boundaries of physics, chemistry and biology to manufacture important, life-saving medicines, we help them by introducing digital tools that make their day-today jobs easier, enabling data-driven decisions.” AI can also enhance diagnosis by analysing individual clinical and operational data, improving hospital administration, providing insights into population-wide health issues from the analysis of demographic data, and improving the manufacture of drugs and other healthcare products. Possibly one of the most ambitious AI and data analytics projects in healthcare has been underway in the UK for a number of years. It is being undertaken through a partnership between chipmaker NVIDIA, The King’s College London Medical Imaging and AI Centre for Value Based Healthcare (AI4VBH) and Owkin, a French American start-up that specialises in AI and federated learning for medical research. They built a network across 15 UK hospitals and three universities that gathers anonymised patient data to identify where clinical practice can be improved, identifying routes to advance research in a host of key therapeutic areas, including neurodegenerative disease, cardiovascular conditions and cancer. Sebastien Ourselin, Professor of Healthcare Engineering at Kings College London, said at the launch the predictive models developed from patient data would be representative and unbiased because they would be trained on the widest possible population of patient data, which in phase one represented one third of the London population. “We truly see this architecture as the future of healthcare informatics.”
Tackling COVID-19 with AI Today, one of the most notable applications of AI has been during the COVID-19 pandemic. Suppression of the virus requires effective and rapid contact tracing, which AI can significantly impact. For example, in September 2020 the Victorian Government trialled AI to assist with contact tracing, adapting a
70
product originally developed for the Australian Defence Force to pick up patterns that might not necessarily be obvious. Alfred Health and Deakin University developed CovidCare, a system that uses AI to remotely manage and triage patients with COVID-19 while they self-isolate at home. It relies on an intelligent decision support system developed by the Applied Artificial Intelligence Institute (A²I²) at Deakin University, and is expected to also be useful in supporting people with chronic diseases. The TIBCO GatherSmart solution uses AI and ML to safely to assess the readiness of personnel to return to the workplace. They can report their symptoms on a daily basis and receive guidance on whether to work from home or go into the office. And to aid research into COVID-19, Amazon Web Services launched CORD-19 Search, a search website powered by machine learning to help researchers quickly and easily search tens of thousands of research papers and documents using natural language questions.
Curbing the rising cost of healthcare Aside from any health benefits that will be gained through the application of AI in healthcare, these applications, and the efficiencies AI can produce, will be needed to prevent healthcare from consuming an ever-greater percentage of Australia’s GDP. Between 2000–2001 and 2017–2018, total spending on health increased in real terms (after adjusting for inflation) from $91 billion to $185 billion—an average growth rate of 4.3 percent per year. According to the Australian Institute of Health and Welfare, health expenditure increased from 7.6 of GDP in 1997–1998 to 10.3 percent in 2015–2016 and declined slightly to 10.0 percent in 2017–2018. That decline is not expected to continue. The Federal Government’s 2021 Intergenerational Report also predicts per person health expenditure will rise from $3250 per year in 2018−2019 to $3970 in 2031−2032. Funding for public hospitals is projected to be the fastest growing component of government health expenditure, nearly doubling between 2020−2021 and 2031−2032.
