Related Experiment Videos
Application of artificial neural networks to establish a predictive mortality risk model in children admitted to a
1Department of Paediatrics, Kwong Wah Hospital, Waterloo Road, Hong Kong SAR, China.
Insights
Artificial neural networks (ANN) show promise in improving paediatric mortality prediction models (MPM). An ANN model demonstrated higher accuracy than the Paediatric Index of Mortality (PIM) in predicting mortality in a paediatric intensive care unit.
Area of Science:
- Artificial Intelligence in Medicine
- Machine Learning for Healthcare
- Predictive Analytics in Pediatrics
Background:
- Paediatric Risk of Mortality (PRISM) and Paediatric Index of Mortality (PIM) are established mortality prediction models (MPM) for paediatric intensive care units (PICU).
- Existing MPMs may have limitations in accuracy for predicting outcomes in critically ill children.
- Artificial intelligence (AI), specifically artificial neural networks (ANN), offers a potential avenue for developing more precise MPMs.
Purpose of the Study:
- To develop and validate an artificial neural network (ANN) model for predicting mortality in a paediatric intensive care unit (PICU).
- To compare the performance of the developed ANN model against the Paediatric Index of Mortality (PIM).
Main Methods:
- A retrospective case series design was employed, analyzing data from patients under 17 years admitted to a Hong Kong PICU (April 2001-December 2004).
- Patient data were randomly allocated to training and validation sets to construct and test ANN models.
- Model accuracy was assessed using the area under the receiver operator characteristics (ROC) curve and calibration, comparing ANN models with PIM.
Main Results:
- Three ANN models (ANN8, ANN9, ANN23) were developed. ANN9, incorporating pre-ICU intubation, demonstrated a high ROC curve (0.957).
- The Paediatric Index of Mortality (PIM) also showed a high ROC curve (0.808), but ANN9 outperformed it in predictive accuracy.
- ANN8 required fewer variables (five) than PIM (eight) for risk calculation, suggesting potential for simplified data input.
Conclusions:
- Artificial neural networks (ANN) can be effectively utilized for developing predictive mortality risk models in paediatric intensive care.
- The ANN model (ANN9) showed superior performance compared to the Paediatric Index of Mortality (PIM) in this cohort.
- Further multicentre studies are recommended to establish a generalizable ANN-based mortality prediction model for diverse PICU settings.
Introduction:
Paediatric risk of mortality and paediatric index of mortality (PIM) are the commonly-used mortality prediction models (MPM) in children admitted to paediatric intensive care unit (PICU). The current study was undertaken to develop a better MPM using artificial neural network, a domain of artificial intelligence.
Methods:
The purpose of this retrospective case series was to compare an artificial neural network (ANN) model and PIM with the observed mortality in a cohort of patients admitted to a five-bed PICU in a Hong Kong non-teaching general hospital. The patients were under the age of 17 years and admitted to our PICU from April 2001 to December 2004. Data were collected from each patient admitted to our PICU. All data were randomly allocated to either the training or validation set. The data from the training set were used to construct a series of ANN models. The data from the validation set were used to validate the ANN and PIM models. The accuracy of ANN models and PIM was assessed by area under the receiver operator characteristics (ROC) curve and calibration.
Results:
All data were randomly allocated to either the training (n=274) or validation set (n=273). Three ANN models were developed using the data from the training set, namely ANN8 (trained with variables required for PIM), ANN9 (trained with variables required for PIM and pre-ICU intubation) and ANN23 (trained with variables required for ANN9 and 14 principal ICU diagnoses). Three ANN models and PIM were used to predict mortality in the validation set. We found that PIM and ANN9 had a high ROC curve (PIM: 0.808, 95 percent confidence interval 0.552 to 1.000, ANN9: 0.957, 95 percent confidence interval 0.915 to 1.000), whereas ANN8 and ANN23 gave a suboptimal area under the ROC curve. ANN8 required only five variables for the calculation of risk, compared with eight for PIM.
Conclusion:
The current study demonstrated the process of predictive mortality risk model development using ANN. Further multicentre studies are required to produce a representative ANN-based mortality prediction model for use in different PICUs.