Artificial Intelligence and Neural Network Based Maternal and Fetal Health Risk Level Prediction and Sensitivity Analysis During Pregnancy

Abstract

Women’s health throughout pregnancy, childbirth, and the postpartum period is referred to as maternal health. Even though most pregnancies and births are successful, issues can occasionally arise and can be extremely painful for mothers and babies. Through a better und erstanding of risk factors and the implementation of earlier and more suitable interventions, predictive modeling can enhance outcomes and assist gynecologists in providing more effective care. The dataset used for analysis was collected from local hospitals in Bangladesh. For the analysis, the main risk factors considered are age, blood pressure (systolic), blood pressure (diastolic), body temperature, maternal heart rate, blood glucose, hepatitis B, Thyroid-Stimulating Hormone (TSH), Serum Glutamic-Pyruvic Transaminase (SGPT), serum uric acid, fetal heart rate, amount of amniotic fluid, fetal movement, and obesity. The dataset contains details of these features for women during pregnancy. Data preprocessing involved encoding categorical values into numeric form, handling missing values through imputation, selecting relevant features, and feature scaling, among other tasks. The data were then divided into training and testing sets, and the model with the best accuracy was determined. Multinomial Logistic Regression (MLR), Naïve Bayes(NB), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), and Multilayer Perceptron (MLP) Neural Network are among the models used. The MLR, Multinomial Naïve Bayes (MNB), KNN, SVM, DT, RF, and MLP algorithms achieved accuracies of 89.58%, 80.13%, 100.0%, 90.23%, 100.0%, 100.0%, and 100.0%, respectively. The DT, RF, and MLP algorithms also yielded the best precision, recall, F1-score, and Receiver Operating Characteristic-Area Under the Curve (ROC-AUC) score, all at 100.0%. The DT, RF, and MLP algorithms performed best overall in terms of accuracy, precision, recall, F1-score, and ROC-AUC score compared to the other algorithms. The input and target variables were also subjected to sensitivity analysis to determine which input parameters have the most significant impact on risk. According to the sensitivity analysis, the following factors significantly impact the risk to the mother’s and fetus’s health: blood pressure (systolic and diastolic), fetal movement, fetal heart rate, and the amount of amniotic fluid. This study will help improve healthcare for expectant mothers and their fetuses by reducing maternal and child morbidity during pregnancy in Bangladesh.