Logo Lanfrica

Donmaston09/mamacare_app

Domain:

healthcare

Record type:

software
Creator:
Don
Host:
MamaCare CAD is an offline-first Progressive Web App designed for rural midwives in low-resource settings. It combines local clinical protocols with optional AI-assisted insights to instantly triage maternal health risks, analyze fetal biometry, and track patient sessions locally. # MamaCare CAD ### Offline-first Maternal Health Computer-Aided Diagnosis Tool A lightweight Progressive Web App (PWA) for midwives and general practitioners at rural clinics in Sub-Saharan Africa and other low-resource settings. Works fully offline after first load. Installs to Android/iOS home screen like a native app. --- ## What it does | Module | Offline | AI-enhanced (online) | |---|---|---| | Fetal biometry analysis (BPD, HC, AC, FL Z-scores) | ✅ | ✅ | | Risk triage (FHR, AFI, BP, presentation, placenta) | ✅ | ✅ | | WHO danger signs checklist (14 signs) | ✅ | — | | Session log — 100 patients, local storage only | ✅ | — | | Biometry reference tables (Hadlock/WHO) | ✅ | — | | AI clinical narrative (Claude Sonnet) | — | ✅ | --- ## Deploy to Netlify (2 minutes) ### Option A — Drag and drop 1. Go to app.netlify.com 2. Drag this entire folder onto the deploy zone 3. Done — live in ~30 seconds ### Option B — CLI ```bash npm install -g netlify-cli netlify login netlify deploy --prod --dir . ``` ### Option C — GitHub 1. Push this folder to a GitHub repository 2. In Netlify: New site → Import from Git → select repo 3. Build command: *(leave blank)* · Publish directory: `.` --- ## Using the AI features The app calls the **Google Gemini API** (`gemini-1.5-flash`) for AI clinical summaries. Because this is an offline-first, client-only application, the API key is handled directly in the browser to maintain the user's privacy and avoid server round-trips: 1. Go to the **Reference** tab in the app. 2. Enter your **Google Gemini API Key** in the "About & Settings" section. 3. The key is securely saved in your browser's local storage and is only transmitted directly to Google's API when generating a clinical summary. No backend proxy or serverless function is required. --- ## Offline behaviour After the first visit, the service worker caches: - `index.html` - `offline.html` - `manifest.json` - Icons All clinical logic (biometry Z-scores, AFI/FHR/BP threshold …