Approved Research
Transfer Learning for Genome Analysis and Personalised Healthcare Recommendation
Lay summary
The idea of personalised or precision medicine is an emerging concept that seeks to address the high variability in each individual and their healthcare requirements. Due to the increasing availability and accuracy of technologies such as DNA sequencing, imaging, and health monitoring, the idea of personalised medicine is a becoming a more feasible and urgent need in the healthcare sector. Having the ability to generate personalised insights for each individual patient can be instrumental in tackling the rising issues in healthcare such as the growing ageing population and the increased prevalence of chronic conditions.
Machine learning (ML) is a subset of artificial intelligence (AI) which continues to "learn" and becomes more accurate in its algorithm with further training. ML is an important tool since it can process the large amount of complex genomic data from whole human genome sequencing and convert it to a usable outcome that can be used to support decision making in healthcare. Therefore, this project aims to use whole genome sequencing data in advanced ML techniques to improve the accuracy, adaptability, and comprehensiveness of health characteristic predictions and provide personalised recommendations for healthcare service and disease prevention. We aim to finishing this project in three years to allow sufficient time to train, test and validate new algorithms and models.
The deliverables of this project will be focused on two sub-studies. These are for stroke and seizure, and for women's health such as endometriosis respectively. By using these two case studies, our team will highlight different areas of healthcare that these algorithms can service effectively. We will deliver novel ML methodologies that will provide predictions based on genome analysis that distils and transfers useful knowledge from multiple sources into an Australian genome analysis model. We also aim to develop a system that will profile individuals and generate personalised healthcare recommendations based on our findings.
The outcome of our findings will generate unique insights on disease prevention which can then be further researched and developed by healthcare scientists. By using ML predictive techniques to provide personalised recommendations, we can aid in offering better and more effective care for patients in our healthcare system. Furthermore, the models and methods developed in this project could also be used for ML tasks in other fields of research that have complicated datasets and require sophisticated methods to extract useful insights.