Machine Learning-Driven Early Detection of Preeclampsia Using Random Forest and Particle Swarm Optimization
JOIV : International Journal on Informatics Visualization · 31.05.2026 · 10 (3) · 1024
DOI: 10.62527/joiv.10.3.4140 ↗Kurz gefasst
Fachartikel in der Zeitschrift JOIV : International Journal on Informatics Visualization, online erschienen am 31.05.2026. Verfasst von R. Topan Aditya Rahman und 5 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
The main purpose of this study is the early prediction of preeclampsia, a serious and complex global medical issue affecting about 5-8% of pregnancies and significantly contributing to maternal and fetal complications. Currently, no widely adopted active screening method exists for early detection, making preventive efforts suboptimal and leading to higher morbidity, mortality, and healthcare burdens. This retrospective study analyzed de-identified medical records of 504 pregnant women at Ansari Saleh General Hospital, Banjarmasin, South Kalimantan, in 2024, focusing on clinical features. Data pre-processing included handling missing values, feature scaling, selection to reduce overfitting, and analysis using the Random Forest ensemble algorithm optimized by PSO for hyperparameter tuning. The optimized Random Forest model achieved an accuracy of 96.08%, an AUC-ROC of 0.971, a precision of 97.06%, a sensitivity of 97.06%, and a specificity of 94.12%. Specifically, the accuracy improved by 1.39 percentage points. These strong results enable reliable prediction of preeclampsia in the first trimester, facilitating timely interventions. Adoption of technology-based health solutions depends on service quality, usability, and clinical workflow integration; thus, developing a user-friendly machine-learning prototype app is highly recommended to empower healthcare workers to improve prenatal care and pregnancy monitoring, thereby reducing the incidence of preeclampsia. For future research, studies should prioritize multi-center, longitudinal datasets from diverse ethnic groups to improve generalizability, integrate real-time physiological data from wearable devices for dynamic risk assessment, and compare hybrid Random Forest-PSO models with advanced deep learning architectures such as transformers or graph neural networks to achieve higher predictive accuracy across diverse global populations.
Zitieren
APA
Rahman, R. T. A., Yuandari, E., Haryono, I. A., Hidayah, N., Iswandari, N. D., & Hateriah, S. (2026). Machine Learning-Driven Early Detection of Preeclampsia Using Random Forest and Particle Swarm Optimization. JOIV : International Journal on Informatics Visualization, 10(3), 1024. https://doi.org/10.62527/joiv.10.3.4140
Vancouver
Rahman RTA, Yuandari E, Haryono IA, Hidayah N, Iswandari ND, Hateriah S. Machine Learning-Driven Early Detection of Preeclampsia Using Random Forest and Particle Swarm Optimization. JOIV : International Journal on Informatics Visualization. 2026;10(3):1024. doi:10.62527/joiv.10.3.4140
BibTeX
@article{Rahman2026,
author = {Rahman, R. Topan Aditya and Yuandari, Esti and Haryono, Ika Avrilina and Hidayah, Nurul and Iswandari, Novita Dewi and Hateriah, Siti},
title = {Machine Learning-Driven Early Detection of Preeclampsia Using Random Forest and Particle Swarm Optimization},
journal = {JOIV : International Journal on Informatics Visualization},
year = {2026},
volume = {10},
number = {3},
pages = {1024},
doi = {10.62527/joiv.10.3.4140}
}
EUVIMED