Intensive Care Unit Mortality Prediction: An Improved Patient-Specific Stacking Ensemble Model

Abstract

The intensive care unit (ICU) admits the most seriously ill patients requiring extensive monitoring. Early ICU mortality prediction is crucial for identifying patients who are at great risk of dying and for providing suitable interventions to save their lives. Accordingly, early prediction of patients at high mortality risk will enable their provision of appropriate and timely medical services. Although various severity scores and machine-learning models have recently been developed for early mortality prediction, such prediction remains challenging. This paper proposes a novel stacking ensemble approach to predict the mortality of ICU patients. Our approach is more accurate and medically intuitive compared to the literature work. Data were prepared and feature selection was processed under the supervision of the ICU domain expert. The data were split into six modalities based on the expert's decisions. For the prediction process, a separate classifier was selected for each modality based on the performance of the classifiers. We utilized the most popular and diverse classifiers in the literature, including linear discriminant analysis, decision tree (DT), multilayer perceptron, k-nearest neighbor, and logistic regression (LR). Then, a stacking ensemble classifier was constructed and optimized based on the fusion of these five classifier decisions. The framework was evaluated using 10,664 patients from the medical information mart for intensive care (MIMIC III) benchmark dataset. To predict patient mortality, extensive experiments were conducted using the patients' time series data of different lengths. For each patient, the first 6, 12, and 24 hours of the first stay were tested. The results indicate that our model outperformed the state-of-the-art approaches in terms of accuracy (94.4%), F1 score (93.7%), precision (96.4%), recall (91.1%), and area under the receiver operator characteristic (ROC) curve (93.3%). These results demonstrate the ability and efficiency of our approach to predict ICU mortality.

Mobile Health in Remote Patient Monitoring for Chronic Diseases: Principles, Trends, and Challenges

“…Managing and integrating the massive data extracted during patient monitoring are considered a daunting task. To take full advantage of the extracted data, various data mining and knowledge extraction tools should be developed to have deep insights into these data to improve knowledge outcomes and decrease costs [ 76 ].…”

“…This system used high volume cloud resources to serve many patients, reduce the hospital and predict early risks. Others [ 76 ] provided CDSS that could predict mortality based on the data of the first 24 h. In [ 77 ], the authors provided a telemonitoring system for diabetes designed to automatically evaluate patient glycemic data that uploaded according to an embedded CDSS. Despite the importance of this contribution, but it did not provide a complete view of a patient’s medical history.…”

Feature Explanations in Recurrent Neural Networks for Predicting Risk of Mortality in Intensive Care Patients

“…Several studies employed descriptive statistics to calculate simple summaries (e.g., minimum, mean, maximum) for the clinical measurements made within the time window (e.g., 24/48 h) and classical ML techniques (e.g., logistic regression and decision trees) to develop a model [ 14 , 15 , 16 ]. Awad et al [ 14 ] used descriptive statistics of the measurements taken during the first 6 h after admission and developed a tree-based method for mortality prediction in ICU patients with an AUC of 0.82.…”

“…This has led to the development of various data-driven approaches for EWSs. Previous studies have used descriptive statistics on clinical observations and tree-based algorithms for mortality prediction [ 14 , 15 , 16 ]. Several studies employed time-series techniques to process real-valued clinical observations for mortality prediction [ 17 , 18 , 19 , 20 , 21 ].…”

Comprehensive Survey of Using Machine Learning in the COVID-19 Pandemic

“…AI is used to tracking patients through smart devices, such as mobile phones, cameras, and other wearable sensors [ 198 , 199 ]. These devices could be used for diagnosing, screening, and continuous monitoring [ 200 ]. Based on data aggregated from these devices, AI could provide useful information for the decision-making process, such as prioritizing the need for respiratory support as well as intensive care unit (ICU) admission [ 58 , 201 ].…”