These national technology networks provide infrastructure that will support a wide range of experimental medicine studies across the UK.
Led by John Gallacher, University of Oxford
Enables rapid access to cohort data, including imaging and genetics.
Stem cells network
Led by Richard Wade-Martins, University of Oxford
Enables work on cellular reprogramming, cellular phenotyping, neuronal physiology and functional genetics.
Led by Paul Matthews, Imperial College London
Enables multi-site molecular and structural imaging studies (such as Deep and Frequent Phenotyping).
The MR-PET harmonisation study
Led by Karl Herholz, University of Manchester
MR-PET is an exciting new scanning technique in which a magnetic resonance (MR) scan and a positron emission tomography (PET) scan are performed simultaneously.
There are currently seven UK-based MR-PET scanners, located at the Universities of Edinburgh, Newcastle, Cambridge, Manchester, and Imperial, King’s and University Colleges London. This study aims to measure scanning variability at and between each of these sites so any differences can be minimised and to allow standardisation of scanning methods. This will be important for future multi-centre studies.
The anonymised imaging data collected will form a reference dataset that can be shared across the DPUK network, and which can be used to evaluate new methods of analysing imaging data.
Led by Clare Mackay, University of Oxford
Imaging data is both large and complex, and is therefore typically stored separately from other cohort data in specialist centres. Until recently, the infrastructure didn’t exist to bring imaging data together at scale, which is important for multi-cohort studies.
DPUK’s imaging informatics provides the infrastructure to bring imaging data together for dementia-related cohorts and multi-centre studies. This is essential for cross-cohort analysis, because different cohorts will have different ways of naming and categorising data. We also provide an environment in which analysts can access the data remotely, rather than having to download extremely large datasets to conduct their analyses. This lowers the technical barrier to performing image analysis for large studies.