Developing an artificial intelligence based prototype model for early prediction of preeclampsia
1
Department of Public Health and Informatics, Bangladesh Medical University, Dhaka, Bangladesh
2
College of Engineering, Qatar University, Doha, Qatar
3
Department of Fetomaternal Medicine, Bangladesh Medical University, Dhaka, Bangladesh
4
Department of Gynecology & Obstetrics, Dhaka Medical College, Dhaka, Bangladesh
5
Maternal and Child Health Training Institute, Dhaka, Bangladesh
Popul. Med. 2026;8(Supplement Supplement 1):A2355
ABSTRACT
BACKGROUND:
Preeclampsia remains a leading cause of maternal morbidity and mortality worldwide, particularly in developing countries like Bangladesh (1). In an era of rapid digital advancement, early prediction of preeclampsia can play a major role in improving maternal health outcomes and support progress toward the Sustainable Development Goals (SDGs) (2).
METHODS:
A quantitative study was conducted using both retrospective and prospective data collected from Yonsei University, South Korea, Bangladesh Medical University (BMU), Dhaka Medical College Hospital (DMCH), and the Maternal and Child Health Training Institute (MCHTI), Lalkuthi, to develop a predictive prototype for preeclampsia. Socio-demographic, obstetric, and clinical variables were obtained through structured interviews and medical records. Associations with preeclampsia status were assessed using Chi-square test, Fisher’s exact test, and Rank sum test. Multiple machine learning algorithms were trained and tested, and the best-performing model was selected. Internal and external validation were then performed to assess model robustness.
RESULTS:
The international dataset, having better structure and completeness, resulted in stronger model training and higher predictive performance, while local datasets were limited by fewer variables and substantial missing data. Among the 249 pregnant women included prospectively, more than half were anemic. 21.7% had current urinary tract infection (UTI), with a higher prevalence among pre-eclamptic women. Socio-demographic factors (age, education, residence), obstetric history (family or prior history of preeclampsia/eclampsia/HELLP syndrome, menstrual problems before conception), and comorbidities (autoimmune disease and hypertension) showed strong associations with preeclampsia. The Extra Trees Classifier demonstrated the best performance (accuracy 98.39%, AUC 99.94%) and achieved 95% accuracy following external validation. Key predictive features included proteinuria, diastolic blood pressure, previous history of preeclampsia, and UTI.
CONCLUSIONS:
The findings uphold the potential of AI-based models for early prediction of preeclampsia. Strengthening health data systems and integrating clinical and socio-demographic information can enhance predictive capacity and contribute to reducing maternal mortality.
Preeclampsia remains a leading cause of maternal morbidity and mortality worldwide, particularly in developing countries like Bangladesh (1). In an era of rapid digital advancement, early prediction of preeclampsia can play a major role in improving maternal health outcomes and support progress toward the Sustainable Development Goals (SDGs) (2).
METHODS:
A quantitative study was conducted using both retrospective and prospective data collected from Yonsei University, South Korea, Bangladesh Medical University (BMU), Dhaka Medical College Hospital (DMCH), and the Maternal and Child Health Training Institute (MCHTI), Lalkuthi, to develop a predictive prototype for preeclampsia. Socio-demographic, obstetric, and clinical variables were obtained through structured interviews and medical records. Associations with preeclampsia status were assessed using Chi-square test, Fisher’s exact test, and Rank sum test. Multiple machine learning algorithms were trained and tested, and the best-performing model was selected. Internal and external validation were then performed to assess model robustness.
RESULTS:
The international dataset, having better structure and completeness, resulted in stronger model training and higher predictive performance, while local datasets were limited by fewer variables and substantial missing data. Among the 249 pregnant women included prospectively, more than half were anemic. 21.7% had current urinary tract infection (UTI), with a higher prevalence among pre-eclamptic women. Socio-demographic factors (age, education, residence), obstetric history (family or prior history of preeclampsia/eclampsia/HELLP syndrome, menstrual problems before conception), and comorbidities (autoimmune disease and hypertension) showed strong associations with preeclampsia. The Extra Trees Classifier demonstrated the best performance (accuracy 98.39%, AUC 99.94%) and achieved 95% accuracy following external validation. Key predictive features included proteinuria, diastolic blood pressure, previous history of preeclampsia, and UTI.
CONCLUSIONS:
The findings uphold the potential of AI-based models for early prediction of preeclampsia. Strengthening health data systems and integrating clinical and socio-demographic information can enhance predictive capacity and contribute to reducing maternal mortality.
| eISSN: | 2654-1459 |
We process personal data collected when visiting the website. The function of obtaining information about users and their behavior is carried out by voluntarily entered information in forms and saving cookies in end devices. Data, including cookies, are used to provide services, improve the user experience and to analyze the traffic in accordance with the Privacy policy. Data are also collected and processed by Google Analytics tool (more).
You can change cookies settings in your browser. Restricted use of cookies in the browser configuration may affect some functionalities of the website.
You can change cookies settings in your browser. Restricted use of cookies in the browser configuration may affect some functionalities of the website.
