Suitable for PhD, Masters or Honours students. This project would suit students with a background in physics, maths, statistics, machine learning, engineering, or a related discipline, with some experience in programming (e.g., in MATLAB).
A major challenge in neonatal intensive care is timely and efficient bedside monitoring of the preterm brain to guide optimal individual care. The overarching aim of this project is to clinically validate novel methods to noninvasively detect acute brain injury and form a prognosis for long-term outcome as early as the first hours after preterm birth. Electroencephalography (EEG) is widely used to monitor preterm brain health, but its diagnostic utility is limited by the need for subjective visual assessments of raw signals or simple trends. These are also prone to the many recording artefacts in intensive care units. We recently developed new metrics for analysing preterm brain activity that enable the detection of injuries and prediction of neurodevelopment, earlier than had been possible before. This project will take the crucial next steps toward taking our new technology to the clinic. This will involve validating and refining our existing metrics using a newly collected, large, multicentre dataset of preterm EEG with full clinical follow-up. There are also numerous technical challenges to solve so that our methods can work smoothly in the real-world intensive care environment. The outcome will be a validated brain monitoring toolbox for neonatal intensive care, ready for immediate implementation in brain monitors.