Design Of An Integrated Early Warning Score (EWS) Application System Using Supervised Learning Model

Read full text Summarise with VIRAT
Authors: Wiwiet Danu Subekti, Hilman Fauzi, Indra Wahyudhin Fathona, Indah Sundari
Cited by: 0
References: 18

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract Early Warning Score (EWS) is a clinical assessment system used to detect early signs of patient deterioration based on vital physiological parameters such as respiratory rate, heart rate, blood pressure, body temperature, and level of consciousness. However, the implementation of EWS in many hospitals is still carried out manually, which can lead to delays in patient management due to limitations in data recording and analysis. This study aims to design and develop an Android-based application that integrates machine learning methods to support the classification and prediction of patients’ clinical risk conditions. The main objectives include: designing a machine learning-based application to assist medical personnel in determining early diagnoses based on physiological data; implementing the Random Forest algorithm as a supervised learning model for automatic EWS calculation; and analyzing the performance and effectiveness of the algorithm in supporting clinical risk assessment. In addition to using standard physiological parameters in EWS, the application expands the scope of assessment by incorporating laboratory and radiology (X-ray) data to provide a more comprehensive risk score for organ failure and mortality. The evaluation results show that the Random Forest algorithm achieved an accuracy of 85.00, precision of 87.50, recall of 85.00, and F1-score of 82.47, with an AUROC value of 0.7733, indicating good classification performance. The model also demonstrated high efficiency with a training time of 0.1340 seconds and a prediction time of 0.0189 seconds. These findings indicate that the model can provide accurate and rapid risk predictions, highlighting its potential as a clinical decision support system in intensive care units.

Fields of Study & Subjects

Artificial Intelligence in Healthcare Machine Learning in Healthcare Healthcare Technology and Patient Monitoring

Publication details

Journal / source
Journal of Physics Conference Series
Volume
3188
Issue
1
Pages
012011–012011
Publication date
1 March 2026
Language
EN