Principal Investigator: Christine Guo
Department: QIMR Berghofer Medical Research Institute
Collaborating lead –
Dr Pierrick Bourgeat – CSIRO, Herston, Australia
Dr Lei Du – Northwestern Polytechnical University, Xi’an, ChinaTags: 27483, body, Brain, genetics, Imaging, lifestyle, MRI
The health of the brain and the body are highly interrelated. Comorbidities of the nervous and cardiovascular systems are commonly reported: cardiovascular changes are often observed in neuropsychiatric disorders and cognitive decline is prevalent in cardiovascular diseases such as diabetes. However, the underlying physiological mechanisms of this interaction remain elusive.
We aim to:
1) improve methodologies for analysing large imaging datasets of the brain and body;
2) characterise the relationship between measures of brain, obesity and cardiovascular health, under the influence of genetics, metabolic and lifestyle factors;
3) prospectively investigate imaging markers that reflect disease risk.
Understanding the interacting pathophysiological mechanisms between the imaging markers and risk factors of neuropsychiatric and cardiovascular diseases could lead to better diagnosis, prevention, and treatment. In the future, when ageing-related diseases develop in some of the cohort, this could be used to investigate how the patterns learnt from the current dataset estimated disease risk and if it could predict disease onset. Therefore, our aims are well aligned with the objectives of UK Biobank project. We will develop improved analytical methods to extract multimodal features of the brain and the body (cortical thickness, brain connectivity, fat distribution, etc). Bayesian models and machine learning algorithms will be applied to identify the correlation patterns across these key measures of brain and body health. Genomic data, mental health and lifestyle factors will be included in the model. An additional aim is to examine the relationship between polygenic risk score of dementia and measures of brain and cardiovascular health to develop imaging markers of disease risk, in combination with dataset from our Australian prospective imaging study. We require the subset of UK Biobank cohort subjects with complete phenotype, genotype information and imaging data – the actual number to be advised by the Biobank. We also request access to their follow-up clinical assessments available in the future.
Project extension – January 2020
We aim to:
4) Using statistics and machine learning approaches for association and prediction analysis to identify genomic interactions as well as the interaction between genomics and disease risk or other health data.