paulmakarimayi/AI-Healthcare-Early-Disease-Detection-Platform-v2

AI-powered early disease screening platform localized for Zimbabwe. Predicts risk for Diabetes, Heart Disease, Dengue, Malaria, Cholera and Rabies using explainable AI (SHAP + WHO-based scoring). Supports English, chiShona and isiNdebele. Includes Doctor/Nurse/Admin accounts and real local hospital contacts. Academic project. # 🩺 AI Healthcare Early Disease Detection Platform — Zimbabwe Edition **Academic Project — Midlands State University** *Built by Paul Makarimayi* A screening-support tool for communities with limited access to doctors, localized for Zimbabwe. Users answer simple symptom/health questions and receive a risk score, an explanation of *why* that score was given, and guidance on where to go next — including real local hospital/clinic contact details. > ⚠️ **This is a screening/referral tool, not a diagnostic tool.** It never > tells a user they "have" a disease — only their risk level and where to > go next. Always consult a qualified healthcare professional. 🔗 **Live Demo:** _[add your Streamlit Cloud link here once deployed]_ --- ## ✨ Features - 📊 **Risk scoring** for Diabetes, Heart Disease, Dengue, Malaria, Cholera, and Rabies - 🔍 **Explainable AI** — SHAP for the ML models, transparent WHO-informed weighted scoring for the rule-based conditions, so every prediction shows *why* - 🏥 **Zimbabwe hospital/clinic recommendations** with real contact numbers (Parirenyatwa Group of Hospitals, Harare General/Sally Mugabe Central Hospital, Avondale Clinic, Arundel Hospital, Nkayi District Hospital) plus national emergency numbers (999 / 994) - 🌐 **Multilingual support** — English, chiShona, and isiNdebele - 👥 **Staff accounts** for Doctor, Nurse, and Admin roles: - Doctors/Nurses request an account → pending until reviewed - Admin approves/rejects requests, monitors login status, and can delegate/change a user's role between Doctor and Nurse - 👨‍⚕️ **Clinician Dashboard** — Doctors/Nurses see all screenings submitted so far --- ## 📁 Project Structure ``` rotahack/ ├── app.py # Main Streamlit application ├── train_models.py # Trains the Diabetes + Heart Disease ML models ├── rule_engine.py # Dengue/Malaria weighted symptom scoring ├── auth.py # Staff accounts: signup, login, admin approval, role delegation ├── tran …