Predicting C-Section Outcomes Among Bangladeshi Women: A Comparative Study of Machine Learning Techniques
Paper
Paper/Presentation Title | Predicting C-Section Outcomes Among Bangladeshi Women: A Comparative Study of Machine Learning Techniques |
---|---|
Presentation Type | Paper |
Authors | Haque, Faisal, Nahar, Sabikun, Haque, Umme Marzia, Hossain, Zakir and Kabir, Enamul |
Journal or Proceedings Title | Proceedings of the13th International Conference on Health Information Science (HIS 2025) |
Journal Citation | 15336, pp. 164-173 |
Number of Pages | 10 |
Year | 2025 |
Publisher | Springer |
Place of Publication | Singapore |
ISBN | 9789819655960 |
9789819655977 | |
Digital Object Identifier (DOI) | https://doi.org/10.1007/978-981-96-5597-7_15 |
Web Address (URL) of Paper | https://link.springer.com/chapter/10.1007/978-981-96-5597-7_15 |
Web Address (URL) of Conference Proceedings | https://link.springer.com/book/10.1007/978-981-96-5597-7 |
Conference/Event | 13th International Conference on Health Information Science (HIS 2024) |
Event Details | 13th International Conference on Health Information Science (HIS 2024) Parent International Conference on Health Information Science (HIS) Delivery Online Event Date 08 to end of 10 Dec 2024 Event Location Hong Kong, China |
Abstract | Background: In recent decades, the rate of caesarean section (C-section) has been rapidly increasing in Bangladesh. This study explores different machine learning techniques to find out potential determinants of C-section in the country. Methods: Data on C-section from 4932 women were extracted from the clustered Bangladesh Demographic Health Survey (BDHS) 2017–18. Chi-square test was applied for feature selection. The machine learning techniques applied included decision tree, random forest, K-nearest neighbors, Gaussian Naïve-Bayes, logistic regression, and support vector machine. The performance of these techniques was evaluated using confusion matrices and receiver operating characteristics (ROC) curves. Results: The prevalence of C-section was found to be 33.0% (1628 out of 4932). Age of mother, wealth index, working status of mother, birth order, number of antenatal care visits, body mass index, birth weight, education, media access, birth interval, place of residence, and administrative division were significant features for the prediction of C-section. Among the ML models, the random forest (accuracy = 0.8143, precision = 0.7907, sensitivity = 0.7977, specificity = 0.8477, F1-score = 0.7939, and area under curve: AUC = 0.90) was found to be the best choice for predicting the C-section among Bangladeshi women. Conclusion: Study findings may help government and non-government organizations, health professionals, and policy-makers to take necessary actions in reducing unnecessary C-section childbirth. |
Keywords | C-section; machine learning; AUC; ROC; random forest |
Contains Sensitive Content | Does not contain sensitive content |
ANZSRC Field of Research 2020 | 420606. Social determinants of health |
420308. Health informatics and information systems | |
Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
Series | Lecture Notes in Computer Science |
Byline Affiliations | East-West University, Bangladesh |
School of Sciences | |
University of Dhaka, Bangladesh |
https://research.usq.edu.au/item/zy5y7/predicting-c-section-outcomes-among-bangladeshi-women-a-comparative-study-of-machine-learning-techniques
13
total views2
total downloads11
views this month2
downloads this month