PhD in Digital Health: Analysing densely captured sensor data for disease prediction

We are offering an exciting PhD opportunity for candidates interested in developing innovative digital health approaches to improve health outcomes. The successful candidate will benefit from a highly multidisciplinary supervisory team and research training environment in the MRC Integrative Epidemiology Unit.

This 4-year studentship is ideal for a talented computer scientist wishing to develop strong interdisciplinary skills at the interface of Computer Science and Digital Health. We offer a fully-funded studentship (starting October 2018). This will be based in the MRC Integrative Epidemiology Unit at the University of Bristol, a leading centre for research excellence in population health science. The successful candidate will have access to an excellent training portfolio of short courses and transferable skills training and be part of a cross-disciplinary cohort of PhD students.

The successful applicant will be supervised by Dr Andy Skinner, Dr Louise AC Millard, Prof Kate Tilling and Prof Deborah A Lawlor. If you are interested please get in touch with one of us for an informal chat.

Summary

Novel wearable technologies such as smartwatches and sensors that continually assess biomarker data, together with more sparsely repeatedly assessed multi-‘omic data provide an opportunity to build models that more accurately predict who will have diseases or adverse health outcomes and who will remain healthy. To date, there has been limited research on the use of such data in healthy populations.

A central aim of this PhD is to determine the extent to which adverse cardio-metabolic health outcomes (e.g. diabetes, hypertension and heart attacks) and adverse pregnancy and perinatal outcomes (e.g. gestational diabetes) can be predicted using data from wearable devices (e.g. continuously measured glucose monitors) in populations who are healthy / women who are pregnant and not known to have any health problems or complications of pregnancy. As part of this aim you will explore the extent to which the wearables improve prediction over and above existing clinical prediction tools and the addition of multiple ‘omics data (genomics, epigenomics, metabolomics) that are available in our cohort studies, often repeatedly assessed.

You will develop and apply cutting-edge prediction methods using both sparse (e.g. every one- or two- years) and dense (e.g. every minute) “time-series” data in healthy people (women and men) and/or healthy pregnant women and evaluate the extent to which such models can accurately identify people’s risk of adverse cardio-metabolic outcomes and/or women’s risk of adverse pregnancy and perinatal outcomes.

This is a flexible PhD where you will also be encouraged to develop and pursue your own research interests. Possible directions include:
(i) exploring the value of adding multiple ‘omics data to currently existing clinical tools for predicting cardiovascular disease and for predicting type 2 diabetes
(ii) exploring the value of digital health monitors that can be used to continuously measure multiple metabolites at the same time.
(iii) the development and feasibility testing of smartphone and smartwatch apps for unobtrusively measuring patterns of behaviours such as eating, drinking and smoking.
(iv) methods for simultaneously analysing several sets of continuously assessed data, multiple ‘omics data repeatedly assessed over long time periods and clinical data.

You will use data from birth cohorts, including the world-leading Avon Longitudinal Study of Parents and Children (ALSPAC) cohort, Born in Bradford family of cohorts, and UK Biobank.

If you are interested please get in touch with one of us for an informal chat. Our email links can be found at the top of this advert.