Respiratory diseases affect hundreds of millions of people in both the economically developing and developed world causing morbidity and mortality . Wheeze disorders (preschool wheeze, viral induced wheeze, asthma) present a major portion of respiratory illnesses with significant socio-economic burden. Current practice of assessing disease severity is principally based on a subjective clinical assessment, which is heavily dependent on training and experience. This leads to significant variance amongst trained professionals when assessing disease severity and consequently variance in healthcare resource use and patient outcomes . Providing technologies for objective clinical decision making could give greater standardisation of healthcare and reduce the burden of training, particularly in low resource settings.
Pulse oxygen saturation monitoring (displaying oxygen saturation (SpO2) and heart rate) is used to support triage decisions in both primary (e.g. refer to secondary care) and secondary care (e.g. discharge from or admit to hospital). The technology is ubiquitous in developed healthcare and in a rapid expansion phase in developing healthcare. SpO2 as displayed has only modest sensitivity and specificity for patient outcomes. Within the high frequency waveform data from which displayed SpO2 are derived are inherent signals that could provide greater objectivity to outputs. Early work in this area aligns with this perspective . We propose that analysis of signals inherent within the pulse wave data (using signal processing techniques) could provide additional objective measures which fused with routinely collected patient data (by leveraging machine learning techniques) has the potential to support better objective patient assessment and reduce variance and risk in care. Such technology would be of value both as a triage tool to help identify patients at greater risk deterioration (physiological instability) warranting admission, and also to support decision making for children in hospital by identifying clinical stability and therefore the clinical confidence to wean bronchodilator and enable faster discharge home.
We will utilise wheeze disorders (preschool wheeze, viral induced wheeze, asthma) as an exemplar to demonstrate the potential of this technology. Wheeze disorders are extremely common in children, represent a significant health and socio-economic burden and have a spectrum of severity which is assessed subjectively by clinicians, resulting in important variance in patient care, healthcare utilisation and outcomes.
We will apply signal processing methods on pulse wave data to identify key signal features and leverage machine learning techniques to fuse those features with routinely collected patient data. The aim will be to differentiate clinical instability in children with wheeze (i.e. triage assessment tool) and also differentiate children where clinical stability identifies an opportunity for objective weaning of medication. Our proposed population is children accessing acute clinical care with pulse oxygen saturation and routine clinical data available. Downloaded SpO2 waveforms would be subject to (a) Signal Processing: Time-domain methods (e.g. peak detection), and parametric (e.g. Autoregressive Modelling)/non parametric (e.g. Fourier Transform) spectrum analysis to discover additional objective measures (b) Machine Learning: Supervised learning algorithms (both linear and non-linear classifiers) for data fusion. We have significant prior expertise in extracting features from time series signals including features such as respiratory rate extracted from SpO2 waveforms .
This project will equip students with advanced computational skills in signal processing, and machine learning applied in healthcare (respiratory illnesses in paediatrics). In the process, student will also get thorough experience of working on a highly interdisciplinary project, where the project spans from literature review to collecting novel data in routine clinical practice to advanced computational techniques thereby providing the student with excellent research training in a project that sits at the interface of machine learning and medicine.
This MRC programme is joint between the Universities of Edinburgh and Glasgow. You will be registered at the host institution of the primary supervisor detailed in your project selection.
All applications should be made via the University of Edinburgh, irrespective of project location. For those applying to a University of Glasgow project, your application along with any supporting documents will be shared with University of Glasgow. http://www.ed.ac.uk/studying/postgraduate/degrees/index.php?r=site/view&id=919
Please note, you must apply to one of the projects and you must contact the primary supervisor prior to making your application. Additional information on the application process is available from the link above.
For more information about Precision Medicine visit: http://www.ed.ac.uk/usher/precision-medicine
Start: September 2020
Qualifications criteria: Applicants applying for a MRC DTP in Precision Medicine studentship must have obtained, or will soon obtain, a first or upper-second class UK honours degree or equivalent non-UK qualification, in an appropriate science/technology area.
Residence criteria: The MRC DTP in Precision Medicine grant provides tuition fees and stipend of at least £15,009 (RCUK rate 2019/20) for UK and EU nationals that meet all required eligibility criteria.
Full eligibility details are available: View Website
Enquiries regarding programme: [email protected]
 T. Ferkol and D. Schraufnagel, “The global burden of respiratory disease,” Ann. Am. Thorac. Soc., vol. 11, no. 3, pp. 404–406, 2014.
 T. A. Florin et al., “Reliability of examination findings in suspected community-acquired pneumonia,” Pediatrics, vol. 140, no. 3, p. e20170310, 2017.
 A. Brandwein, K. Patel, M. Kline, P. Silver, and S. Gangadharan, “Using Pleth variability as a triage tool for children with obstructive airway disease in a pediatric emergency department,” Pediatr. Emerg. Care, vol. 34, no. 10, pp. 702–705, 2018.
 S. A. Shah, S. Fleming, M. Thompson, and L. Tarassenko, “Respiratory rate estimation during triage of children in hospitals,” J. Med. Eng. Technol., vol. 39, no. 8, pp. 514–524, 2015.