Project ID BE-MI2026_10

ThemeBE-MI

Co Supervisor 1A Dr Michele Orini Faculty of Life Sciences & Medicine, School of Biomedical Engineering & Imaging Sciences, Digital Twins for HealthcareEmail

Co Supervisor 1B Dr Sara White Faculty of Life Sciences & Medicine, School of Life Course & Population Sciences, Department of Women & Children’s HealthEmail

Developing computational methods to improve pregnancy outcomes using wearables

“Abnormal fetal growth affects more than 20,000 women in the UK every year, and it can lead to significant neonatal and offspring complications, potentially impacting health across the lifespan. Multiple drivers of abnormal fetal growth have been considered, including maternal nutrition, physical activity, comorbidity and genetics, but their impact on pregnancy outcomes remains poorly understood. The overarching aim of this project is to determine what drives abnormal fetal growth and to propose solutions to optimise fetal health and development.
This project will use unique data from a large study recruiting pregnant women of White, Black and South Asian descent. A unique aspect of this project is the depths and breadth of data collected during pregnancy, which includes detailed clinical (multiple blood, stool and urine samples, fetal ultrasound, etc.) and socioeconomical data, as well as being enriched by continuous data streams from wearable devices, including accelerometers and continuous glucose monitors. This provides the exciting opportunity to objectively measure physical activity, sleep, blood glucose level and their interaction, which are thought to be key drivers of abnormal fetal growth, but whose role is still poorly characterised. A key gap that this project will address is the development and optimization or signal processing and AI methods to transform these complex, multi-dimensional and multi-modal data into accurate and precise predictors of fetal growth. Our student will develop a combination of clinical expertise and computational skills (signal processing, machine learning and artificial intelligence) which are essential to advance digital health strategies. The student will be supported by a supervisory team with complementary expertise in both clinical and data science.
The PhD workplan for a 4-year PhD will be as follows:
Year1: Develop and validate algorithm for assessment of physical activity (e.g. time spent in sedentary, moderate and vigorous activity and number of steps per day) and sleep (e.g. total sleep duration and sleep fragmentation) from wearable accelerometers.
Year2: Implement existing and develop novel approaches to analyse continuous glucose monitor data and to identify unique patterns related to pregnancy.
Year3: Develop computational models to describe the complex interaction between physical activity, sleep and blood glucose variability in pregnant women.
Year4: Develop statistical and machine learning models to predict abnormal fetal growth and major newborn outcomes.
For a 3-year PhD, development work from year-1 and year-3 detailed above will be streamlined in the first 2 years.

3-months rotation project: Assessing the association between physical activity and blood glucose level in pregnancy using existing algorithms. These algorithms derive basic metrics of both physical activity and blood glucose variability that can be used as a starting point for the project.

Representative Publications

1) Jamieson et al, Accuracy of smartwatches for the remote assessment of exercise capacity, Scientific Reports, 2024. https://doi.org/10.1038/s41598-024-74140-x 2) Orini et al, Premature atrial and ventricular contractions detected on wearable-format ECGs and prediction of cardiovascular events, European Heart Journal – Digital Health, 2023. doi: 10.1093/ehjdh/ztad007 3) Bhatt et al, Validation of a popular consumer-grade cuffless blood pressure device for continuous 24 h monitoring, European Heart Journal-Digital Health. https://doi.org/10.1093/ehjdh/ztaf044

2) Gunabalasingam et al, Interventions in women with type 2 diabetes mellitus in the pre-pregnancy, pregnancy and postpartum periods to optimise care and health outcomes: A systematic review; Diabetic Medicine, 2025. 10.1111/dme.15474 2) Kolozali, S et al, Explainable early prediction of gestational diabetes biomarkers by combining medical background and wearable devices: a pilot study with a cohort group in South Africa. 2024; Journal of Biomedical and Health informatics. 10.1109/JBHI.2024.3361505 3) White SL, et al. Rational Testing: Screening and Diagnosis of Gestational Diabetes. BMJ, 2023; https://www.bmj.com/content/381/bmj-2022-071920