ademide-star/-african-neurohealth-dashboard

This diagnostic tool was developed with a strong emphasis on contextual and cultural relevance for African populations # 🧠 Stroke & Memory Loss Diagnostic Tool This diagnostic tool was developed with a strong emphasis on contextual and cultural relevance for African populations. Beyond conventional biomedical variables, it incorporates lifestyle, environmental, and psychosocial stressors that reflect lived realities such as noise pollution, use of herbal treatments, and sleep quality. To further enhance specificity, the system optionally captures users’ cultural group (e.g., Yoruba, Hausa, Swahili's) to allow future modeling of ethnoculturally-informed health patterns. These efforts aim to bridge the cultural gap in AI-driven health diagnostics, ensuring more equitable and representative digital health solutions across Africa. An AI-powered web and mobile diagnostic tool designed to predict individual risk for stroke and memory loss using clinical, behavioral, and environmental factors relevant to African populations. ## 🌍 Project Goals - Predict stroke and memory decline using real-life health, lifestyle, and environmental indicators - Provide a free, scalable tool to aid early intervention in resource-limited settings - Educate users with actionable lifestyle and wellness suggestions ## ✅ Features - Stroke and memory loss risk prediction using machine learning models - Custom stress scale and MMSE estimator built into dashboard - Geolocation and noise exposure tracking for context-aware results - Data stored securely in Supabase backend - Deployed via Streamlit (web + APK compatible) ## 📈 Model Performance - **Stroke Prediction:** Accuracy = 86.7%, ROC AUC = 0.9649 - **Memory Loss Prediction:** Accuracy = 93.3%, ROC AUC = 0.9907 ## 💻 Technologies Used - Python, Streamlit, scikit-learn, joblib - Supabase (backend data store) - Mobile-ready via APK wrapper ## 🔐 Author & Credits Developed by **Adebimpe John Omolola Olamide** Supervised by **Prof. Bamidele Owoyele** (University of Ilorin) and **Prof. Owolabi Mayowa** (University of Ibadan) Supported by Growing Data-science …