Identifying New Biomarkers of Parkinson's From Routine Brain Imaging

We intend to use historical CT/MRI/nuclear medicine brain scans to identify novel imaging biomarkers of prodromal Parkinson's disease. The primary data source for the study will be MRI and CT brain scans, whilst nuclear medicine imaging brain imaging (DAT scans) will be used to validate models produced and provide a functional outcome measure of brain dopamine uptake. We shall utilise a an artificial intelligence approach to compare scans of PD cases with matched controls in order to identify these imaging biomarkers.

A list of participants with a diagnosis will be compiled. This list, together with relevant clinical data, will be linked with historical CT/MRI/nuclear medicine scans carried out over the preceding years. A control group of matched non-PD scans will also be compiled.

The dataset will be anonymised and a bespoke ML pipeline will be used to identify imaging featureswhich may be indicative of prodromal PD. This initial stage will be carried out at University Hospital Plymouth NHS Trust (UHPNT), the Royal Cornwall Hospital NHS Trust (RCHNT) and Cornwall Partnership NHS Trust (CPNT). If successful, findings will be validated in a larger sample of scans compiled from hospitals regionally and nationally.