Transfer Learning for Genome Analysis and Personalised Recommendations
Using AI to improve health predictions
Research Leads: Distinguished Professor Jie Lu, Professor Yi Zhang, Associate Professor Guangquan Zhang
Collaborators: 23Strands
Led by Distinguished Professor Jie Lu in partnership with 23Strands, this ARC Linkage project developed advanced AI and transfer learning methods to improve genomic risk prediction and personalised healthcare recommendations, with applications including stroke, infertility and women's health.
This project aims to improve the accuracy, adaptability, and comprehensiveness of health characteristic predictions and provide personalised recommendations for healthcare service and disease prevention.
This project capitalises on Australia’s strengths in science and technology to generate wider economic benefits through improved knowledge translation and commercialisation, and partnerships with industry. It is aligned with the National Science and Research Priority: Health, to have “better models of health care and services that improve outcomes, improved prediction, identification, tracking, prevention and management of emerging local and regional health threats”.
The project will revolutionise our understanding of the associations between genes and diseases by developing AI empowered techniques and tools for knowledge transfer, predicting correlations and providing evidence-based recommendations, allowing healthcare professionals to build efficient applications that integrate the frontier of genomics into treatment regimens and better models of health care services for disease prevention. A target area of application is chronic diseases which accounts for 36% of the total burden of disease in Australia and 38% of this cost burden is preventable.
The deliverables from the project include uncertainty learning and multi-source transfer learning methodologies for predictions based on genome analysis that distils and transfers useful knowledge from multiple sources into an Australian genome analysis model.
A federated cross-domain recommender system will be developed to profile individuals and generate personalised recommendations. The outcomes are expected to create a paradigm shift in learning-based prediction and personalised recommendations to support healthcare services in complex environments.
AI for Endometriosis: From Genomics to Personalised Management
Personalised care for endometriosis patients
Research Leads: Distinguished Professor Jie Lu, Dr Hua Lin (23Strands), Dr Kairui Guo, Dr Mengjia Wu
Collaborators: 23Strands, Digital Health Cooperative Research Centre
A collaborative project involving UTS, 23Strands and the Digital Health Cooperative Research Centre, focusing on AI-driven genomic analysis and personalised management of endometriosis. The project develops transferable AI models and integrates genomic and clinical information to support risk prediction, disease characterisation and personalised healthcare applications.
Under Distinguished Professor Jie Lu and a dedicated multidisciplinary team, is using fuzzy transfer learning to build explainable AI models that successfully translate global datasets into a multi-modal Australian Endometriosis dataset.
Read more about Endometriosis Awareness Month from Digital Health CRC.
Technical Intelligence for Gene Research and Healthcare
Intelligent bibliometrics for tracking and predicting technological change
Research Leads: Distinguished Professor Jie Lu, Dr Yi Zhang
AAII Research Lab: Decision Systems and e-Service Intelligence
Collaborators: 23Strands
The rapid development of gene research, together with the significantly increase of related scientific documents, challenges entrepreneurs in systematically capturing knowledge with limited expert support and identifying business opportunities in a short time.
Intelligent bibliometrics provide an express solution of constructing a knowledge base for a given gene-related disease. This project developed a systematic toolkit to analyse large-scale scientific documents for technical intelligence e.g., identifying key players of a given domain, profiling key research areas and their associations, and tracing the evolutionary pathways of research topics over time.
Representation Learning on Large Scale Administrative Data
Improving policy with data insights
Research Lead: Dr Guodong Long
AAII Research Lab: Data Science and Knowledge Discovery
Collaborators: Australian Commonwealth Department of Health
The key to effective policy development is implementing the right policy for the right group of people at the right time. To support this ethos, the public service is keen to develop predictive modelling methods to leverage the information hidden within its administrative data about different population cohorts.
In a large-scale service system, such as the Australian healthcare system, user data exists in complex structures reflective of the complexity of human interactions. Appropriate ways to model this complex structured data is a challenge common to many organisations, from sector-scale commercial services to national-scale public services.
A generalised representation of complex data is the most desirable data format for processing, particularly for large organisations and government agencies that typically deal with data on a massive scale. In particular, utilising powerful deep learning tools could be one of the implementation framework for representation learning on large-scale administrative dataset. The framework will preserve the complex relationships in an unsupervised or self-supervised manner, and to protect privacy in a decentralised training scheme without direct access of the sensitive data.
The DSKD Lab has collaborated with Australian Commonwealth Department of Health a series of projects that apply the most advanced machine learning techniques to model and analyse healthcare administrative data to promote effective policy development.
Using Virtual Reality to Create Certainty in Clinical Decisions using Complex Genomics
Helping doctors make better cancer treatment decisions
Research Leads: Professor Paul Kennedy, Associate Professor Dan Catchpoole
AAII Research Lab: Biomedical Data Science
Collaborators: The Children’s Hospital at Westmead, Western Sydney University, UTS Animal Logic Academy, Samurai Punk, Sony Foundation, Tour de Cure
The complex human genome underpins the biological mechanisms of a patient’s cancer. Presently the knowledge provided by the genome is inaccessible to clinicians making treatment decisions for patients.
We develop machine learning and data analytics methods to model patients in a virtual reality space emphasising appropriate genomic similarities and differences. Our VR technology allows cancer specialists and analysts to move into the 3D space and to explore the patient cohort from within. It enables clinicians to find patients with genomic similarities with other sufferers, moving far beyond simple statistical clustering methods to enable clinicians to uncover genomic relationships, and inform decision making for treatment regimens with a breadth and accuracy that has not been previously possible.
Clinicians can explore the three dimensional space of the patient cohort with data simplified into clusters of genomic similarities, for clinicians to make more informed decisions of treatment regimes for individual patients.