Artificial Intelligence & Machine Learning to Predict Exercise Adherence in Cancer Survivors

Clinical evidence highlights the significant health and well-being benefits of physical activity for cancer survivors during and following cancer treatment. However, cancer-specific barriers often result in cancer survivors participating in less physical activity than their healthy counterparts. Several psychosocial and physical determinants of physical activity participation have been identified in this population, yet how these, in isolation or in combination, influence patients’ adherence to specific exercise prescription over time is not clearly understood. This may, at least in part, be because conventional statistical methods are unable to appropriately determine the associations and relationships among studies with these multiple parameters.

The aim of this PhD program of research is to develop a new clinical data science machine learning system that utilises the psychosocial and physical determinants of physical activity to improve our understanding of exercise adherence in cancer survivors. We expect that the improved understanding of adherence through this approach may provide a future platform for researchers and clinicians to appropriately prescribe individualised exercise interventions based on the best likelihood of adherence. Ultimately, this would result in enhanced health outcomes for people living with and beyond cancer.