Artificial Intelligence for non-invasive hypoglycaemia detection and prediction using wearable sensors

PhD Scholarship:

Artificial Intelligence for non-invasive hypoglycaemia detection and prediction using wearable sensors

School of Engineering, University of Warwick

University Hospitals Coventry and Warwickshire NHS Trust

Application deadline: Open, subject to the position being filled

Start date: October 2021*, *flexible

Duration: 3 years

The project: This is a joint venture between the School of Engineering at the University of Warwick and the Human Metabolic Research Unit at the University Hospitals Coventry and Warwickshire (UHCW). This project will provide a unique opportunity to apply artificial intelligence (mainly deep learning) to biomedical signal processing. The candidate will join a multidiscipline team of biomedical engineers, computer scientists and clinicians having the opportunity to participate proactively through the full lifecycle of the project. This will include the experimental design, data-driven systems modelling using machine learning approaches; system validation via experimentation in a state-of-the-art metabolic research unit; ultimately leading to automatic detection and prediction of significant events (e.g., hypo- or hyperglycaemia) in real-life and in the health care setting.

In the past 3 years, we developed an AI model detecting hypoglycaemia via ECG in healthy subjects. In order to develop this technology, we enrolled two cohorts of healthy participants, which underwent the following protocols:

• 25 participants (over 65 years old) were continuously monitored for up to 36 hours in the Human Metabolism Research Unit (HMRU) of the UHCW (paper here https://doi.org/10.1016/j.bspc.2020.102054 ).
• 8 participants (mean age 38 years, range 26-58 years) were continuously monitored for 2 consecutive weeks while conducting their regular daily activities (paper here https://www.nature.com/articles/s41598-019-56927-5)

Both cohorts wore non-invasive sensors for continuous electrocardiogram (ECG) recording and Continuous Glucose Monitors (CGMs) for continuous blood glucose measurements. Our AI models detected nocturnal hypoglycaemic events achieving 90% sensitivity and specificity in controlled environments, and 87.5% sensitivity and 81.7% specificity in real-life scenarios.

The candidate will have access to the above data and models and will support the research team in exploring how the same approach can work for diabetic patients. The current study aims at recruiting up to 60 patients at the HMRU.

Eligibility: UK resident candidates with a 1^st or 2.1 UK Honours degree or equivalent in Engineering, Mathematics, Physics, Computer Science, or Systems Biology will find this studentship especially relevant. Applications from candidates with a medical science background who are interested in artificial intelligence and biomedical time series analysis are welcome too. The candidate will be supported in deepening advanced methods and tools for deep learning. However, excellent data analysis and computing skills are essential (e.g. fluency in programming in Python, Java/C/C#, MATLAB; familiarity with tools and libraries for deep-learning) as well as an ability to work in a clinical environment, liaising with a range of healthcare professionals and patients is a must as a significant amount of time will be spent in the research facility at the UHCW. We are looking for confident, self-motivated candidates with good interpersonal skills, and abilities to perform excellent research leading to high-quality research outcomes and publications.

Funding: The studentship covers 100% fees (UK level only*) and a standard RCUK level stipend of £15,609.*Fees statement

Supervision and mentoring: the candidate will have two supervisors Dr John Hattersley, Director of the Human Metabolic Research Unit at the University Hospitals Coventry and Warwickshire (UHCW), and Dr Leandro Pecchia, Director of the Applied Biomedical Signal Processing and Intelligent eHealth Lab and the School of Engineering, University of Warwick.

How to apply: Applicants should send a cover letter (2 pg max) outlining motivation and suitability for this project, and a full CV to Dr Leandro Pecchia l.pecchia@warwick.ac.uk and Dr Hattersley John.hattersley@warwick.ac.uk

If you are successful at the interview, you will be required to fulfil the entry requirements set by the University of Warwick. As soon as you have a University ID number you will be invited to upload your degree certificate, transcripts, and a personal statement that explains your specific research interests and why you should be considered for this award. Details available here www.warwick.ac.uk/pgrengineering.

Application Form Course Details:

Department: School of Engineering

Course Type: Postgraduate Research

Course: MPhil/PhD in Engineering (P-H1Q2)

Equality, Diversity and Inclusion

The University of Warwick provides an inclusive working and learning environment, recognising and respecting every individual’s differences. We welcome applications from individuals who identify with any of the protected characteristics defined by the Equality Act 2010.