PhD Studentship in Mechanistic, Physiological Models of Intensive Care Patients and Classification via Machine Learning

Location of position: London
Duration of Studentship: 3 years from start date
Annual stipend: £17,280

Vacancy Information:
This is an exciting opportunity since we are looking for a PhD studentship funded by the Department of Mechanical Engineering at UCL as part of CHIMERA project.

CHIMERA is one of four national Maths in Healthcare Hubs announced in March 2020. CHIMERA, which stands for Collaborative Healthcare Innovation through Mathematics, EngineeRing and AI, is a multidisciplinary Hub which brings together experts in mathematics, statistics, data science and machine learning, with unique, high volume and rich data sets from both adult and paediatric Intensive Care Units provided through embedded Project Partnerships with Great Ormond Street Hospital and University College London Hospitals.
Studentship Description
We will explore available data and review current literature as well as public-domain packages, e.g. to build mechanistic, physiological models relevant to ICU patients, focusing on breathing and cardiac performance.

During the project, we will then carefully consider model assumptions and we will analyse structural uncertainties for these models. We will then proceed to generate a ‘cohort’ of candidate models of varying complexity/structure for validation and selection. These models will be parametrised and validated using data already existing in partner hospitals.

For the best performing candidate models, much larger cohorts of virtual subjects will be established by varying the values of the clinically interesting parameters, overcoming the limitations of smaller cohorts. We will implement deep-learning architectures to classify patients using each candidate model and identify patients at greatest risk of a given event.

The at- risk patient types will be analysed further and compared via a similarity score. The clinicians will be explicitly consulted to see how/which of the risk classifications obtained might best assist clinical decision making.

Person Specification:
- Have achieved (or are predicted) a first class or upper second class honours undergraduate degree (or equivalent international qualifications or experience).An MSc is also preferred, though not essential.
- Our preferred subject areas are Physical Sciences (Computer Science, Engineering, Mathematics and Physics) or any core Engineering discipline (e.g. Bioengineering/Biomedical Engineering, Mechanical Engineering, Chemical, Electrical Engineering, etc.). All applicants must be able to demonstrate strong mathematical skills.
- Strong programming skills in Matlab, Python, C++, Java or any other programming language.
- Experience in physiological modelling would be a plus.
- Applicants whose first language is not English are usually required to provide evidence of proficiency in English by UCL. Further details can be found on the following UCL web page


Eligibility:
UK/EU applicants

If you meet the requirements set above you can apply directly by e-mail to Professor Vanessa Diaz ([Email Address Removed]) with the following information:

- A recent CV

- The full transcript of exam results (listing all subjects and their corresponding grades/marks)

- A cover letter stating how this opportunity meets your research interests.

Individuals in their final year of study should list all modules/grades for which the results are already known.