Principal Investigator: Dr Christopher Long
Siemens HealthCare GmbH, Erlangen, GermanyTags: 36681, causal inference, Machine Learning, matched-sampling, observational-studies
According to the World Health Organisation (2012), neurological disease in its various forms afflicts tens of millions of people worldwide and in some of its domains is forecast to rise significantly. In prevalence studies of Alzheimers Disease (AD) for example, figures are predicted to rise from around 36 million today to over 100 million by 2050. While there are limited treatment options available for many of these afflictions, it is likely that future treatment strategies will be most effective if applied at the earlier stages of disease.
It is the aim of this one year project to increase the clinical utility of large-scale neuroimaging datasets through improved statistical modelling of underlying disease factors as they relate to neurological disease. With the ultimate goal of developing a statistical framework for assessing different treatment regimens, we seek both to enhance early disease detection and advance understanding of the complex neurological mechanisms that foreshadow disease onset.
Project extension March 2019:
Observational neuroimaging studies undertaken at-scale can offer deep insight into the complexities underlying neurological and psychiatric disease. While neuroimaging studies have shown promise over the last decade in the investigation, diagnosis and in treatment monitoring of these disorders, causal inference methodology in population neuroimaging is relatively underdeveloped compared to areas where large datasets are common as, for example in epidemiological studies and the social sciences.
The primary aim of this project is to develop and extend statistical and algorithmic techniques from the fields of Causal Inference and Machine Learning to applications in observational neuroimaging studies. Where possible we seek to build statistical models that go beyond capturing simple associational models to richer classes of technique capable of providing definitive causal links between interventions or risk factors, nascent imaging markers of disease and health outcomes.
Our main clinical focus will be the identification of cohort subjects most at risk of dementia within a timeframe that allows manipulable interventions (pharmacological, or lifestyle/dietary recommendations) to be more effectively applied. In the case that a decision is made to treat and acquire follow-up post-treatment data on these subsets, we will use the developed techniques to assess the relative causal efficacy of the prescribed interventions.
We would like to extend the scope of the project to use genetic data because this extension would allow us to take advantage of ideas in Instrumental Variable Analysis (and Mendelian Randomisation) in characterising these causal relationships relating imaging to disease risk. In particular we would like to combine Mendelian Randomisation techniques with machine learning to allow the use of large neuroimaging datasets as exposures or mediating variables in our causal analyses.
Last updated Mar 12, 2019