We are developing artificial intelligence based prediction models, to identify risk of serious infections and sepsis in children, in real time.
Artificial Intelligence enabled SEPsis TRacking ONline
The Prediction Model
We are developing groundbreaking algorithms by applying machine learning and statistical methods to routinely collected electronic health records, to create our prediction models.
Real-Time Risk Management
Our model will then be integrated with standard software being used in emergency departments. The model will track a patient's journey from triage, monitoring vital signs, triage notes, repeat observations and blood tests to make accurate predictions of sepsis risk, in real-time.
Why We Do It
Sepsis is a leading cause of global and hospital mortality without a gold standard diagnostic test available to date. Early identification and timely administration of antibiotics improves a patient's outcome.
Consensus definitions for paediatric sepsis are often difficult to apply to children attending emergency departments. Furthermore, screening tools in current clinical practice, have demonstrated limited accuracy and reliability to identify onset of sepsis.
We have taken a new approach, using a novel method for early recognition of sepsis risk.
Why Use
Machine Learning?
Machine Learning (a subset of artificial intelligence) has a superior ability to predict patient outcomes compared to traditional approaches.
The advantages of machine learning approaches include its ability to process complex nonlinear relationships between predictors and yield more stable predictions.
