Early Prediction of Preeclampsia by Using Ensemble Machine Learning With MM-XAI Approach

Abstract

Preeclampsia is a pregnancy complication that endangers both the mother and fetus, so early detection is necessary to prevent serious complications. This study employs mixed methods which combines quantitative and qualitative approaches to design a preeclampsia risk prediction system for pregnant women. The system utilizes four machine learning algorithms: Logistic Regression, Decision Tree, Support Vector Machine (SVM), and Random Forest. We conducted the evaluation process by using a Multi-Method Explainable AI (MM-XAI) approach with Shapley Additive Explanations (SHAP), Local Interpretable Model-agnostic Explanations (LIME), and Permutation Feature Importance (PFI) to enhance the transparency and ease of interpretation of the results. Data were obtained from the Anak Air Community Health Center in the Koto Tangah district of Padang City. Clinical variables included systolic and diastolic blood pressure, blood sugar, body temperature, heart rate, age, and urine protein. The results showed that Random Forest and SVM achieved the highest accuracy at 78%, with consistent predictions across all risk categories. MM-XAI emphasized blood pressure and blood sugar as the dominant features which influenced preeclampsia risk prediction. MM-XAI achieved high accuracy and made it easier for healthcare professionals to understand the determining factors of the prediction. This supports quick, accurate, and accountable medical decision-making.