Approved research

Development of infrastructure to generate bespoke brain imaging derived phenotypes

Lay summary

There is a wealth of analytical methods for extracting imaging derived phenotypes (IDPs), some of which are specific to individual diseases. We seek to develop IT infrastructure to enable researchers to interface their bespoke analysis pipelines to the UKB data without having to rely on bandwidth- and resource- intense downloads. Working with the MRC Dementias Platform UK (DPUK) we will develop this infrastructure and apply bespoke analysis pipelines in participants who are and are not at increased genetic risk for Alzheimer?s Disease (APOE e4). Although developed for dementia, the infrastructure will be available for all imaging researchers. The proof of principle study will establish the effect of a dementia risk gene on brain structure and function in a larger population than ever performed previously. This proposal will both contribute to, and facilitate future research using brain imaging phenotypes, including, but not exclusive to, dementia research We are setting up infrastructure using technology called XNAT in close collaboration with Alan Young. After UKB data access procedures have been followed, a researcher will bring their UKB 'key' to the XNAT infrastructure and 'see' the data to which they have been given access. They can then run pre-packaged analysis pipelines on the data through the XNAT interface. As a proof of principle we will run hippocampal and white matter lesion segmentation pipelines developed at UCL on UKB participants aged >60 For this proof of principle study we will run pipelines on a subset of participants from the imaging pilot who have undergone brain MRI who are aged >60 (approx 3000), but in the future this infrastructure will be available for all imaging participants.