AI for Prediction of Fluid Bolus Treatment Responses in the Paediatric Intensive Care Unit
Lead Research Organisation:
University College London
Department Name: Institute of Health Informatics
Abstract
Characterising the response to fluid resuscitation in children admitted to a paediatric intensive care unit (PICU) and predicting their response using time-series analysis of bedside monitoring data.
Context of Research - Fluids are administered in children with shock during critical illness, this is partly because they can be given quickly and easily, with an immediate response: an increase in blood pressure and decrease in heart rate. However, there is currently conflicting evidence of whether using fluids can cause an increase in mortality. This research aims to identify and characterise the response to fluid using AI techniques, in order to understand whether a patient's response can be predicted. Using AI to understand the potential risks of administering fluids can have the impact of knowing when fluids should be given, in turn potentially reducing rate of mortality.
Aims and objectives - The aim of the project is to (i) characterise the response to fluid resuscitation in children admitted to a paediatric intensive care unit (PICU) and (ii) predict the response to fluid using bedside monitoring data recorded at 5-second frequency.
The research methodology - segmentation, multi-scale decomposition, and varying clustering techniques of time-series data (dynamic time warping or neural networks). The amplitude and duration of any response to fluids using auto-regression. Deep neural networks for classification of response to fluids.
EPSRC's strategies and research areas - Artificial intelligence technologies
Companies/Collaborators - Great Ormond Street Hospital, and the CHIMERA team at UCL.
Context of Research - Fluids are administered in children with shock during critical illness, this is partly because they can be given quickly and easily, with an immediate response: an increase in blood pressure and decrease in heart rate. However, there is currently conflicting evidence of whether using fluids can cause an increase in mortality. This research aims to identify and characterise the response to fluid using AI techniques, in order to understand whether a patient's response can be predicted. Using AI to understand the potential risks of administering fluids can have the impact of knowing when fluids should be given, in turn potentially reducing rate of mortality.
Aims and objectives - The aim of the project is to (i) characterise the response to fluid resuscitation in children admitted to a paediatric intensive care unit (PICU) and (ii) predict the response to fluid using bedside monitoring data recorded at 5-second frequency.
The research methodology - segmentation, multi-scale decomposition, and varying clustering techniques of time-series data (dynamic time warping or neural networks). The amplitude and duration of any response to fluids using auto-regression. Deep neural networks for classification of response to fluids.
EPSRC's strategies and research areas - Artificial intelligence technologies
Companies/Collaborators - Great Ormond Street Hospital, and the CHIMERA team at UCL.
People |
ORCID iD |
Simon Robert Arridge (Primary Supervisor) | |
James Stainer (Student) |
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/S021612/1 | 01/04/2019 | 30/09/2027 | |||
2421160 | Studentship | EP/S021612/1 | 28/09/2020 | 30/09/2024 | James Stainer |