PhD Studentship: Development of a Clinical HFNC Decision Support System: Using Machine Learning to Help Clinicians Predict the Outcome of HFNC Therapy

An opportunity exists to join the world-leading Interdisciplinary Collaboration in System Medicine (ICSM) Research Group to work on an industrially funded PhD project that will use state-of-the-art machine learning and AI methods to develop clinical decision support systems for high-flow nasal cannula therapy.

The project:

Concerns around the possible negative consequences of delayed escalation of treatment are a significant factor inhibiting the use of High Flow Nasal Cannula Therapy among some clinicians. Using data from clinical collaborators around the world, this project will develop a user-friendly software tool, based on state-of-the-art machine learning models [1], that can predict the outcome of HFNC therapy in individual patients with higher accuracy than current clinical indices, early in their course of treatment [2,3]. This will increase confidence among clinicians that treating a patient with HFNC will lead to a positive outcome, and help them identify earlier patients who are failing on HFNC and need treatment escalation.

Key References:

[1] H. Yu, S. Saffaran, A. Ali, C. Henry, N. Mustfa, A. Thomas, A. Rajhan, S. Isrhad, L. Weaver, R. Tonelli, L.S. Menga, Q. Zhang, M. Einollahzadeh Samadi, A. Schuppert, J.G. Laffey, L. Camporota, A.M. Esquinas, D.L. Grieco, M. Antonelli, L. Martins de Lima, L. Kawano-Dourado, I.S. Maia, A. Biasi Cavalcanti, E. Clini, T.E. Scott, and D.G. Bates, "In-Hospital Testing of NIVPredict - An AI Tool for Early Prediction of Non-Invasive Ventilation Outcome in Acute Respiratory Failure", Critical Care, 2026. pdf
[2] H. Yu, S. Saffaran, R. Tonelli, J.G. Laffey, A.M. Esquinas, L. Martins de Lima, L. Kawano-Dourado, I.S. Maia, A. Biasi Cavalcanti, E. Clini, and D.G. Bates, "Machine learning models compared with current clinical indices to predict the outcome of high flow nasal cannula therapy in acute hypoxemic respiratory failure", Critical Care, 2025. pdf
[3] H. Yu, S. Saffaran, R. Tonelli, J.G. Laffey, Q. Zhang, A.M. Esquinas, L. Martins de Lima, L. Kawano-Dourado, I.S. Maia, A.B. Cavalcanti, E.Clini, and D.G. Bates, "Evaluating the effect of heart and respiratory rate measurement errors on the ability to predict the outcome of high flow nasal cannula therapy: a multi-centre study", Critical Care, 2025. pdf

Scholarship:

The award will cover the UK tuition fee level, plus a tax-free stipend, currently £21,805, paid at the prevailing UKRI rate for 3.5 years of full-time study.

Eligibility:

This studentship is available to home students only.

The candidate should have a good 2.1 Bachelors, or Masters degree in Engineering, Computational or Physical Sciences, or Physiology/Medical Sciences. This project will suit those with a keen interest in mathematical/computational modelling and simulation applied to real-world medical challenges.

How to apply:

Candidates should submit an expression of interest by sending a CV and supporting statement outlining their skills and interests in this research area to https://www.warwick.ac.uk/engineeringscholarships/dgbss1/app. If this initial application is successful, we will invite you to submit a formal application.

Candidates must fulfil the University of Warwick entry criteria and obtain an unconditional offer before commencing enrolment.

Should your application for admission be accepted, you should be aware that notification of acceptance for the PhD does not constitute an offer of financial support. Successful scholarship candidates will receive an official communication from the School of Engineering to confirm their award.