Contextual Framework for Remote Intelligent Monitoring and Detection System for Prediction of Pregnancy Complications

Maternal health disorders can cause complications and harmful incidents in women during pregnancy. To minimize risks, this research developed a platform powered by a supervised machine learning model (SMLM) to support remote intelligent monitoring and prediction of pregnancy complications caused by hypertensive disorder, gestational diabetes, and related indicators. The study used real-world datasets with six UCI Machine Learning Repository features to identify and predict Maternal Health Risk (MHR) factors. A Support Vector Machine (SVM) algorithm was applied to construct the classifier model, which was trained and evaluated using StratifiedKFold cross-validation (k=10). The model achieved 80% accuracy with a precision, recall, and F1-score of 77%. The outcome of this work is the P-Health mobile application, designed to record and track blood pressure, blood sugar, heart rate, body temperature, and weight, while predicting the risk level of pregnancy complications through inference from the integrated machine learning model. Developed with Kotlin and Android Studio, the application enables healthcare practitioners to remotely monitor patients’ vitals in real time. This innovation addresses the challenge of early detection of pregnancy complications and provides continuous monitoring and assessment. The findings suggest that P-Health can improve early detection and timely intervention, helping medical specialists minimize maternal health risks. The system also has the potential to raise public awareness of maternal health issues, thereby contributing to the prevention of complications during pregnancy.