Key words: big imaging data, radiotherapy, prediction model, treatment outcomes and side-effects

"Learning from Big Imaging Data to Predict Radiotherapy Treatment Outcomes and Side-effects"

The prevalence of cancer is an increasing healthcare issue as it is the predominant cause of death worldwide. The growing cancer burden is caused by several factors including population growth, aging, and the changing prevalence of certain causes of cancer during social and economic development. To address the global cancer burden, new technologies, for instance Artificial Intelligence (AI), have been applied in the workflow of cancer care from diagnosis to treatment. For cancer treatment, especially radiotherapy, new innovations are not only useful to provide comprehensive treatment plans, but also able to reduce radiotherapy- induced side-effects which may exist in patients during and (long) after treatment. This thesis focuses on AI-based quantitative imaging techniques (e.g., radiomics) that have the potential to assist doctors and patients to make individualized treatment decisions.