Approved research

Integrated predictive risk models using multimodal imaging and genetics

Lay summary

The stated aim of the UK Biobank is 'improving the prevention, diagnosis, and treatment of a wide range of serious and life-threatening illnesses'. Two key steps towards this goal are the development improved disease prediction models and developing insights that can assist individuals to reduce their risk of diseases. While there is a wealth of research around image analysis and disease prediction from MRI-derived biomarkers and genetics separately, there has been less work around longitudinal models, which leverage whole-body MRI and genetics. The goal of this study is to develop machine learning methods to predict disease progression and risk to assist patients and clinicians in care decisions. To achieve this goal, we plan to develop algorithms which leverage whole-body MRI, genetics, and other phenotypic data. With this rich information, we test the feasibility of and design tools aimed at improving the (1) differentiation of an individual patient's risk, (2) assisting clinicians in directing patients to optimal clinical care, and (3) providing insights for patients that can help them reduce their risk.