Predicting the need for mechanical ventilation in critical care patients with early respiratory failure using machine learning techniques

Lead Research Organisation: University College London
Department Name: Institute of Health Informatics

Abstract

Prognostication is an essential tool for risk adjustment and decision making in the intensive care unit. Even with increasingly large amounts of electronic health record data available, the goal of unravelling complex disease mechanisms to better forecast patient outcomes remains largely unattained in critical care. Current predictive models are typically derived from linear models limited to data from admission or the first 24 hours and do not explicitly handle incremental information from trends. In this project, we aim to develop a non-linear prognostic performance estimation model to predict the risk of intubation of critical care patients at any point during their admission using continuous time-series data along with time independent variables available within the NIHR Health Informatics Collaborative for Critical Care (HIC-CC) dataset, deploying machine learning techniques including data fusion algorithms. In addition, the technical requirements for collecting, storing, extracting and analysing physiological waveform data within HIC-CC will be established. The MRes can be scaled up to a PhD by integrating physiological waveform data into the HIC-CC dataset and test whether prediction capability increases.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/S021612/1 01/04/2019 30/09/2027
2245964 Studentship EP/S021612/1 23/09/2019 30/09/2023 Thilina Nuwan Jayatilleke