Parkinson’s disease has debilitating motor symptoms of tremor in the limbs, slowness of movement, and freezing, unable to move. A highly effective treatment is electrical stimulation deep in the motor regions of the midbrain. But surgery for this deep brain stimulation is only offered to around 2% of all patients, and about a quarter of those who receive it have poor outcomes. Optimising the selection of patients for deep brain stimulation will widen access to treatment, improve treatment outcomes, and prevent harm. The goal of this project is to test how fusing clinical data, neuroimaging, and video assessments could optimise the selection of patients. The project will be in collaboration with MachineMedicine (London), a MedTech company specialising in Parkinson’s disease, and the movement disorders clinical team at St George’s Hospital, London. The goal of the collaboration is to build an app used in-clinic for patient selection. MachineMedicine are leading the app development, building on their existing app for capturing movement video in-clinic. The clinical team at St George’s are running a trial of Parkinson’s patients to acquire the essential clinical data on patient symptoms, neuroimaging (including fMRI of spontaneous brain activity), and video capture of movements. In joining this collaboration, the PhD student will be trained in data-science and machinelearning tools, including how to extract and analyse MRI and fMRI data, in fusing data across modalities, and in developing a machine-learning pipeline for predicting patient outcomes. These predictions will be tested against the 12-month follow-up data from the St George’s trial patients. The student’s further training will include a 3-month placement at MachineMedicine, and visits to St George’s clinic.