Professor Farah Magrabi – Appointment to The Global Partnership on AI
Professor Farah Magrabi has been nominated by the Federal Department of Industry, Science, Energy and Resources, as one of Australia’s representatives for a 3-year term on the Global Partnership for AI (GPAI). Professor Magrabi will be working with the Montreal Centre of Expertise on Responsible AI.
The GPAI secretariat sits within the OECD, with Centres of Expertise based in Montreal and Paris. and aims to foster international collaboration on the responsible use and development of AI. Australia joined the GPAI in June 2020 as one of the founding members, alongside Canada, the European Union, Germany, India, France, Italy, Japan, New Zealand, the Republic of Korea, Singapore, Slovenia, the United Kingdom, and the United States of America.
The objectives of the GPAI are to support and guide the responsible development and use of AI; facilitate international collaboration in a multi-stakeholder manner; and be a global point of reference for multi-disciplinary research on AI.
Membership of the GPAI allows Australian experts to collaborate with international counterparts to build capacity and understanding of the use AI technologies to address national as well as global challenges such as drought, bushfires, energy stability and even pandemics such as COVID-19.
Australia’s other GPAI experts are Professor Elanor Huntington – Future of Work; Professor Genevieve Bell – Responsible use of AI; Dr Paul Dalby – Data Governance; Professor Toby Walsh – Innovation and Commercialisation. Professor Enrico Coiera completes his 3-year membership of the AI and Pandemic Response working group at the end of 2022.
MRFF Research Data Infrastructure grant success
Appropriate Antimicrobial Use: Scaling Surveillance Using Digital Health
A group of researchers from the Doherty Institute, the Peter Macallum Cancer Centre, the University of Melbourne, and RMIT University has received a Research Data Infrastructure MRFF grant ($2,962,654.00) to develop and implement digital health solutions for surveillance of the quality of antimicrobial use in healthcare.
The emergence of the global health challenge of antimicrobial resistance has highlighted a need to promote the safe and judicious use of antimicrobial medications, which is one of the pillars of the National Antimicrobial Resistance Strategy in Australia.
A foundational requirement for antimicrobial stewardship programs in Australia is the ability to monitor the appropriateness of antimicrobial use. This is needed across the healthcare system, including in hospitals, primary care, aged care and animal health settings.
The Appropriate Antimicrobial Use: Scaling Surveillance Using Digital Health project will bring together a multidisciplinary team of antimicrobial stewardship experts from human and animal health with digital health experts to establish a new research platform for surveillance of antimicrobial use across health settings.
University of Melbourne Professor Karin Thursky, director of the National Centre for Antimicrobial Stewardship at the Peter Doherty Institute for Infection and Immunity and director of the Guidance Group at the Royal Melbourne Hospital will lead this project. Prof. Thursky, who is an infectious diseases physician and health services researcher, has played a leading role in developing local and national antimicrobial stewardship initiatives, focusing particularly on the use of digital health technologies.
Professor Thursky says this project will provide an exciting opportunity to collaborate with digital health experts and data scientists to use novel data science methods to facilitate the automated surveillance of antimicrobial prescribing in hospitals, primary care and veterinary practices.
AAAiH’s SQE working group responds to RANZCR ethical principles
Several members of the Safety, Quality and Ethics (SQE) working group collaborated to provide feedback to the Royal Australian and New Zealand College of Radiologists (RANZCR) about their ethical principles to guide the use of AI in radiology. The nine RANZCR principles which were first published in 2019 were revised and open to public consultation in August 2022.
The principles aim to guide:
- The development of standards of practice for research in AI tools
- The regulation of market access for AI tools
- The development of standards of practice for deployment of AI tools in medicine
- The upskilling of medical practitioners in AI tools, and
- The ethical use of AI tools in medicine
This is an ambitious set of aims given the diverse range of stakeholders involved in developing and deploying AI in radiology. However, the principles serve an important function in drawing attention to relevant ethical considerations such as patient safety, data privacy, avoiding bias and attributing responsibility. Like all high-level principles, the challenge lies in putting them into practice.
The SQE submission focused on ways of clarifying the principles to make them more specific and therefore more practical. For example, Principle Four: Transparency and Explainability stipulates that “When designing or implementing an AI tool, consideration must be given to how a result that can impact patient care can be best understood and explained by a discerning medical practitioner.”
The submission asked how this could be realised in practice and what level of explainability is required given that clinicians and patients vary in their capacity to understand and use information. It also encouraged RANZCR to clearly distinguish and define what is meant by ‘transparency’ and ‘explainability’ as these terms are widely used but not always consistently. It was suggested that ‘AI explainability’ be used to refer to a (technical) explanation about how AI generated a specific output; and ‘transparency’ for any information about AI including very general information (e.g. information that AI is being used in a given procedure) to very specific information (algorithmic parameters, training, validation, clinical testing information, etc).
The SQE working group recognise the difficulty in crafting ethical principles and congratulate RANZCR for their leadership; and hope that the submission will provide valuable feedback in this revision process.
Written by Wendy Rogers for the RANZCR ethical principles submission working group of the Safety, Quality and Ethics program.
RANZCR Ethical Principles for Artificial Intelligence in Medicine
International Workshop – Towards Transparent and Explainable AI in Healthcare
On 1 September 2022, Macquarie University Law School, in collaboration with AAAiH and the Australasian Society for Computers and Law (AUSCL), hosted an international workshop exploring issues surrounding the transparency and explainability of ML/AI-based technologies in healthcare. The workshop held industry and academic panel discussions, with a diverse range of speakers from the health, legal and AI sectors. It provided a valuable opportunity for multidisciplinary perspectives and insight on what further research is needed in the field.
Organisations represented included: Australian Institute of Innovation; Macquarie University; University of Surrey, School of Law; UCLA, School of Medicine; KPMG, Ashurst Lawyers; Max Planck Institute for Competition & Innovation, Munich; Chinese University of Hong Kong, Faculty of Law; New South Wales Bar; University of Queensland, Medical School; Consumers Health Forum of Australia; University of Sydney, Chief Investigator with the ARC Centre of Excellence for Automated Decision-Making and Society.
The recording of the hybrid event will soon be available on AUSCL Youtube channel.
Quantum computers offered first to research institutions in 2023
AAAiH member Fujitsu Australia is accepting expressions of interest for quantum computer orders. In August, Fujitsu shared with media outlets the intention to make the computers commercially available to Universities and Research companies. Fujitsu’s longer-term direction is to offer quantum computing on public cloud.
Fujitsu has been working with Riken since last year through the creation of a new research centre called the Riken RQC-Fujitsu Collaboration Centre located in Wako city, Saitama prefecture. A team of 20 researchers work there, combining Riken’s quantum computer technology using superconducting circuits with Fujitsu’s computing technology and knowledge of quantum technology applications. The first quantum computer Fujitsu intends to offer in 2023 has 64 qubits and the intended market is research companies working in the fields of medicines, materials, and financial forecasting.
To register an Expression of Interest please complete the ‘Get In Touch’ details at the bottom of this page Fujitsu IT Solutions and Services for the Healthcare Industry in Australia and New Zealand
Alliance invited to IMDRF/GMTA joint workshop – September 12th
The Alliance was invited by the Therapeutic Goods Alliance (TGA) to the International Medical Device Regulators Forum / Global Medical Technology Alliance (GMTA) joint workshop on Updates and Considerations for Artificial Intelligence Medical Devices. Professor Farah Magrabi gave a presentation about AI technologies in healthcare settings and was on a panel discussing ways to mitigate bias in AI and medical devices. The event was attended by over 200 members of the GMTA, Medical Device Regulators, Medical Device Industry and other bodies.
While Artificial Intelligence (AI) and Machine Learning (ML) have the potential to make a significant difference in health and care settings, more can be done to fully harness the benefits. The workshop gave participants the opportunity to hear from medical device manufacturers, regulators and technologists and other stakeholders on the current and emerging trends impacting this important area, as well as current regulatory approaches and opportunities.
The GMTA represents companies producing the medical devices, diagnostic products and health information systems that are transforming health care through earlier disease detection, less invasive procedures and more effective treatments. The companies represented by GMTA produce nearly 85 percent of the health care technology purchased and utilised annually around the world.
Please tell us your news
The final newsletter for 2022 will be out at the start of December. Please send us details of news items, including forthcoming meetings and events, and notable achievements: info@aihealthalliance.org