Joburu/maternal-health-ai-predictive-analytics

AI-driven remote monitoring and predictive analytics for maternal health in rural Kenya — published research (2024) # Maternal Health AI — Predictive Analytics for Rural Kenya ## Overview This project presents an AI-driven framework for remote monitoring and predictive analytics to enhance maternal health outcomes in rural Kenya, with a focus on high-risk pregnancy identification and early intervention. Based on peer-reviewed publication: **Oburu, J.J. & Simwa, R. (2024).** Localized AI-Driven Remote Monitoring and Predictive Analytics to Enhance Maternal Health in Rural Kenya: Bridging Accessibility and High-Risk Pregnancy Management. June 2024. ## Problem Statement Rural Kenya faces critical gaps in maternal healthcare access. High-risk pregnancies often go undetected until complications arise, contributing to preventable maternal and infant mortality. ## AI Approach - Predictive modelling for high-risk pregnancy classification - Remote patient monitoring system design - Real-time alert framework for rural healthcare providers - Localized model training on Kenya-specific health data ## Key Methods - Machine learning classification models (R & Python) - Logistic regression and ensemble methods - Remote sensing and mobile data integration - DHIS2 health data system compatibility ## Key Outcomes - Framework for scalable AI-assisted maternal care in low-resource settings - Predictive model identifying high-risk pregnancies before critical events - Recommendations for policy integration with Kenya's health systems ## Tools Used - R (caret, ggplot2, randomForest) - Python (scikit-learn, pandas) - DHIS2 integration ## Real-World Impact This research directly informs healthcare policy and digital health strategy for county governments and national health institutions in Kenya. ## Publication [View full paper — June 2024] ## Author **Dr. Jeffar J. Oburu** PhD Applied Statistics | MSc Actuarial Science University Lecturer & AI Research Consultant 📧 jeffaroburu234@gmail.com 🔗 linkedin.com 🔗 github.com