Sustainable IoT-enabled predictive analytics for maternal health risk prediction: A deep learning approach
Year: 2025
Type: article
Abstract: Maternal health is a significant concern, especially in low-resource environments with limited healthcare infrastructure, economic constraints, and access. The rise of the Internet of Things (IoT) and deep learning presents promising solutions. This study explores the deep learning approach to create an IoT-driven predictive analytics model to evaluate maternal health risks. By using the Maternal Health Risk Dataset, the ratio of systolic to diastolic blood pressure was engineered (BP_ratio). The evaluation included random forest, support vector machine, and gradient boosting alongside the deep learning model. The deep learning model achieved a balanced performance with an accuracy of 71.17%, a precision of 72.78%, a recall of 70.29%, and an F1-score of 65.71%. These results suggest that integrating IoT with predictive analytics can enhance early detection and intervention, reducing maternal mortality and morbidity. The study offers practical insights for healthcare stakeholders and policymakers in low-resource environments to implement efficient and scalable healthcare solutions.
Authors: Abayomi Agbeyangi, Jose Lukose
Institution: Walter Sisulu University
Language: English
Subfield: Health Information Management
Field: Health Professions
Domain: Health Sciences
Open Access status: diamond
OpenAlex ID: W4408271989