Predicting post-stroke dementia from CT neuroimaging and other biomarkers

Approximately a third of people who have a stroke will go on to develop dementia. At present, there are no known biomarkers which accurately predict the risk of developing post-stroke dementia. Computer tomography (CT) is a neuroimaging technique commonly performed in the acute phase following a stroke, thus creating a vast amount of stroke patient data. The aim of my PhD is to identify CT biomarkers that predict who will develop post-stroke dementia.

The project will initially start by conducting a systematic review of the literature surrounding CT imaging and post-stroke cognitive decline. I will be aiming to identify the CT biomarkers associated with post-stroke dementia and investigate the link between the CT biomarkers and the cognitive deficits observed in the disease. In order to receive training on how to conduct a high quality systematic review I will also be working on an on-going Cochrane review within the Cochrane Dementia Group, Glasgow. Alongside the systematic review I am aiming to perform an individual participant meta-analysis of the CT biomarkers of post-stroke dementia that I identify.

Following the systematic review and meta-analysis, I am planning to utilise the Karolinska Memory Clinic Cohort data in Sweden. In Sweden, I will join an on-going project and will perform a nested case-control study identifying neuroimaging biomarkers for post-stroke dementia. The study will involve looking at CT and MRI imaging data, cognitive data, CSF data and genetics in dementia patients who have previously had a stroke and comparing it to subjects who haven't had a stroke. The data that I am planning on accessing has been collected from eight different hospitals in Sweden. The neuroimaging data will have been coded and scored by the hospital, however, it is likely to vary in quality and it may be necessary to learn how to read the brain images and independently score them myself. I will be undertaking taught neuroimaging courses as part of the PhD and am planning on observing a neuroimager in order to build upon my knowledge of neuroimaging.

After I have identified potential CT biomarkers of post-stroke dementia I will aim to validate them by accessing data from additional stroke and dementia cohorts in both Scotland and Sweden. This is with the aim to develop a prediction model of post-stroke dementia. There may also be the opportunity to work with existing research groups to use my findings to develop personalised intervention studies, targeting treatments to reduce the risk of post-stroke dementia.