OdaloV/neotrack

Offline AI PWA predicts neonatal deterioration 6-12hrs early from routine vitals. For low-resource Kenyan health facilities. SMS + in-app alerts. Installs on Android tablets. No internet required # Neonatal Early Warning System (NEWS) > AI-powered early warning system for neonatal deterioration in low-resource Kenyan health facilities --- ## Problem Statement **In Kenya, thousands of newborns die from preventable causes each year.** Most health facilities caring for newborns lack continuous monitoring equipment. Nurses check vitals manually and intermittently — every 4-8 hours. By the time deterioration is obvious (hypothermia, sepsis, respiratory distress), it's often too late. ### The 3 Delays Problem: 1. **Delay in detection** - No continuous monitoring 2. **Delay in recognition** - Lack of clinical decision support 3. **Delay in response** - No alert system for nurses ### Key Statistics: - **6,500+** maternal deaths per year in Kenya (many with neonatal impact) - **Preterm birth rate:** 10-12% in low-resource settings - **Neonatal mortality:** 21 deaths per 1,000 live births - **Facilities with continuous monitoring:** <15% in rural areas --- ## Solution **NEWS (Neonatal Early Warning System)** - An AI-powered, offline-first PWA that: 1. **Analyzes routinely collected vitals** (temp, HR, RR, SpO2, weight, gestational age) 2. **Predicts deterioration risk** 6-12 hours in advance using machine learning 3. **Generates real-time alerts** via SMS and in-app notifications 4. **Works offline** in facilities with poor internet connectivity 5. **Installs like an app** on any Android tablet - no app store needed