Develop a web-based system using the Naïve Bayes algorithm to predict asphyxia neonatal

Abstract

Introduction: Most cases of perinatal asphyxia are caused by conditions unrelated to labor. When asphyxia occurs during childbirth, it is usually caused by an obstetric emergency that was not detected during pregnancy. It is essential to prevent asphyxia by identifying the incidence of asphyxia during pregnancy. Several studies have been conducted to identify asphyxia problems developing by predictive models. However, there has been no development of a system for predicting birth asphyxia during pregnancy and carried out in primary health facilities. Purpose: Develop a web-based system using the Naïve Bayes (NB) algorithm to predict asphyxia neonatal using a dataset of antepartum risk factors in primary health facilities. Methods: This study employed research and development, which consists of 4 stages, namely literature study, development stage, expert validity, and trial. Results: A system that health workers in primary health facilities can use to predict asphyxia neonatal and recommend referrals for determining the place of childbirth has been successfully created. The system performance test predicted asphyxia neonatal with all NB evaluation values reaching more than 98%, and the prediction accuracy in the respondent test included in the High Accuracy category (MAPE value 9.06%). Conclusion: The development of a web-based system using the NB algorithm has been proven to be able to predict asphyxia neonatal and can be implemented for health workers as an effort to anticipate delays in handling cases of asphyxia neonatal because of the predicted results along with recommendations for focusing mothers with the risk of babies born asphyxia to find out possible childbirth places.

Keywords

• asphyxia
• asphyxia neonatal prediction
• bayes theorem