EPSRC PhD Studentship
30 July 2018
The School of Public Health has funding available for an Engineering and Physical Sciences Research Council (EPSRC) PhD Studentship to start in October 2018.
Our projects fulfil the EPSRC research aims while delivering the world leading medical research associated with our Faculty. Successful candidates may have a mathematics, physics or engineering background or area of specialism as well as in interest in biomedical sciences.
The studentship will be for 3.5 years, so as to include a funded 6-month writing-up period. An enhanced stipend of £18,000 p.a. is offered, together with a Conference Fund of £300 p.a. and fully funded tuition fees at the home rate. Successful students will become part of the Imperial MRC DTP cohort and benefit from the network and support provided for these students.
Choose a project from those listed below:
Project 1 - A Bayesian hierarchical framework to evaluate policy effects through quasi-experimental designs in a longitudinal setting
Supervisor: Professor Marta Blangiardo
In social sciences, environmental science and public health, the evaluation of the causal effects of policy interventions affecting the lives of large proportions of the population is of crucial importance. These causal effects must be obtained using statistical methods that are rigorous, ensure accountability and produce results that are relevant for future decisions. In this context, randomised controlled trials cannot be carried out for ethical reasons. Thus to answer these questions researchers need to rely on observational studies, making the casual interpretation of the results more challenging.
The main objective of the PhD is to develop a comprehensive, flexible and robust methodology for observational studies to obtain causal inference from longitudinal data in which a policy intervention has taken place. The proposed framework will: i) account for complex hierarchies as well as spatial and temporal dependencies by means of Bayesian models; ii) provide evidence synthesis/data augmentation where negative outcome controls or related controls (potentially on a different scale and collected separately) are incorporated into the analysis; iii) build an exhaustive sensitivity analysis using both simulations and multiple analyses in order to determine amongst other things what data we need to be able to reliably estimate the effects of interest. The proposed methodological framework will be applied to evaluate the impact of the Low Emission Zone (LEZ), implemented in London in 2008 to reduce air pollution from traffic.
Project 2 - Investigating the dynamics of falciparum and vivax malaria elimination
In many parts of the world, malaria elimination — defined by the World Health Organization (WHO) as the absence of locally acquired malaria cases in the country — is being considered as a target because of recent successes in reducing disease burden. This is not the first time that malaria elimination has been considered on a large geographic scale as in 1955 WHO initiated the ‘Global Malaria Eradication Programme’ (GMEP). There are many lessons that can be learnt from the GMEP data that are highly relevant today as this data source has been neglected as it remains on paper in the WHO archive. The student will use previously underexamined data from the GMEP, current national malaria elimination programmes and the published literature, to investigate the epidemiology and monitoring of malaria in near-elimination settings. This work will entail building novel mathematical models of transmission dynamics for falciparum and vivax malaria in very low transmission settings and comparing them to data. The student will aim to elucidate the role of immunity, particularly that provided by the Duffy negative antigen which is known to confer resistance against falciparum malaria. The project will train an individual with a multidisciplinary skillset encompassing mathematical modelling, statistical inference and epidemiological analysis whilst providing valuable insight into the dynamics of disease elimination.
Project 3 - Methods for the rapid analysis and visualisation of poliovirus genomic surveillance data
With the increasing use of genetic sequencing in pathogen outbreak response, methods to analyse genetic sequencing data are increasingly important in epidemiology. In the case of poliovirus, sequence data play a critical role in the global eradication programme, allowing identification of the origins of new outbreaks and distinguishing wild-type from vaccine-derived polioviruses. Recent advances in sequencing technology, including rapid field-based testing that can generate a poliovirus genome within 8 hours of sample collection, offer even greater potential benefit to the programme. This sequencing technology potentially allows rapid identification of transmission routes, gaps in surveillance and the likely extent of circulation. However, there is an urgent need to develop the mathematical and statistical methods and software tools to rapidly analyse and visualise these sequence data. The PhD candidate would develop these approaches using the R statistical programming language, Bayesian phylogenetic methods and online visualisation tools. They would be based within the Vaccine Epidemiology Research Group (VERG) led by Prof Grassly, which is the WHO collaborating institute on polio data analysis and modelling. VERG works closely with the Global Polio Eradication Initiative to analyse global polio surveillance data, including laboratory support for next-generation of poliovirus in stool and environmental samples. The student would benefit from interactions among staff and students in VERG, as well as the MRC Centre for Global Infectious Disease Analysis for which Prof Grassly leads the vaccine priority area.
Project 4 - What are the options and potential for using road charging to improve population health?
Policy and practice partner: Steve Watkins, Director of Public Health at Stockport Council & Transport and Health Study Group Co-chair (Policy)
Aims: The aim of this study is to evaluate how different methods of road charging may impact on population health through changes to travel patterns, physical activity and air pollution levels.
Methods: This research will address three related questions:
What are the impacts of previous road charging schemes (e.g. the London Congestion Charge)? This will be addressed via a systematic review of the available literature plus epidemiological analyses of routinely available health data for U.K. specific interventions
(2) What is the estimated impact of road charging scenarios on travel patterns, physical activity and air pollution levels for different local authority areas? This will be addressed using routinely available environment data (e.g. from Defra’s Automatic Urban and Rural automatic air quality monitoring network) and health outcome data (from the Imperial Small Area Health Statistics Unit) to estimate local air quality and prevalence of health outcomes, along with estimates from objective 1, to predict likely impacts of road charging schemes on these exposures and outcomes.
(3) Are the overall health impacts of road charging different across demographic and socioeconomic subgroups, and do these vary according to local characteristics such as traffic density or available alternatives to car use?
Outcomes: This project will provide novel evidence on the impacts of road charging schemes on population health to inform and support local decision making.
Project 5 - Quantifying children’s exposure to advertising and fast food outlets using imagery with deep learning
Children’s exposure to targeted advertising and outlets selling tobacco and fast food has adverse effects on their risk behaviours and health. Quantifying these exposures dynamically in large scale, and capturing spatial variations (e.g. relating to proximity to schools/playgrounds, areas with different socio-economic characteristics), is challenging because most traditional methods rely on costly surveys. Increasing availability of digital technologies at low cost (e.g. cameras, mobile phones) coupled with advances in deep learning methods can potentially advance surveillance capabilities to quantify these exposures at much larger scales and finer spatial and temporal resolution. The aim of the proposed project is to collect/collate and analyse imagery data from two selected cities to map exposure levels. We will use London and Accra as case studies; a data collection methodology will be devised to gather street level imagery data from cameras mounted on vehicles and crowd-sourced from smart phones. Existing data collection platforms (e.g., Mapillary) will be used to store and map locations of collected images. The project will then use advanced deep learning methods for object detection and segmentation for detecting advertising and outlets from street level imagery to map and quantify exposure levels and variations.
How to apply
- Applications are invited from candidates with a Master’s degree or equivalent in appropriate discipline. Applicants must hold, or expect to obtain, a first or upper second-class undergraduate degree or UK equivalent, in an appropriate subject from a recognised academic institution, and must meet College entry requirements.
- You must select your preferred project from the list of available projects. Please include your preferred project choice in your application. Your application must include a two-page CV and 500-word statement detailing why you would like to apply for this scheme.
- The application will be shortlisted and, if successful, you will be invited to attend an interview by the relevant department and submit a formal application to the College.
Please submit your project choice and application documents to Helen King (firstname.lastname@example.org) by 30 July 2018.
Residence eligibility guidance
Please refer to the EPSRC student eligibility webpages for full details, and see below for the residence eligibility criteria.
To be eligible for a full award a student must fulfil all the following criteria:
- Have settled status in the UK, meaning they have no restrictions on how long they can stay.
- Been ‘ordinarily resident’ in the UK for three years prior to the start of the studentship. This means they must have been normally residing in the UK (apart from temporary or occasional absences).
- Have not been residing in the UK wholly or mainly for the purpose of full-time education (this does not apply to UK or EU nationals).