SehatDori 🪢

Smart Maternal Care Powered by Graph Intelligence

---

🚀 Overview

Sehat Dori is a next-generation maternal healthcare platform designed to save lives in rural areas by intelligently connecting patients, ASHA workers, hospitals, and risk factors using a graph-based approach.

Every year, thousands of maternal deaths occur due to:

• ❌ Missed checkups
• ❌ Poor tracking of patient history
• ❌ Lack of real-time risk identification
• ❌ Fragmented healthcare systems

👉 Sehat Dori solves this using relationship-driven intelligence.

---

UI

---

💡 Key Idea

Instead of storing healthcare data in isolated tables, we model it as a connected graph:

Patient → Visits → Hospital  
       → Risk Factors  
       → ASHA Worker  

This allows us to:

• Detect high-risk pregnancies in real-time
• Enable priority-based care
• Provide actionable insights to field workers

---

🏗️ Tech Stack

⚡ Core Technologies

• Frontend: React + Tailwind CSS + Framer Motion
• Backend: Node.js + Express
• Graph Database: TigerGraph + MongoDB
• Authentication: JWT

---

🔗 Architecture

React App (ASHA Dashboard)
        ↓
Node.js Backend (API Layer)
        ↓
TigerGraph (Risk Engine)
        ↓
Insights → Alerts → Actions 🚨

---

🔥 Features

🚨 Real-Time Risk Detection

• Identifies high-risk pregnancies instantly

• Combines multiple factors:

  • Hemoglobin levels
  • Missed visits
  • Medical conditions

---

📱 Doctor Dashboard

• Simple, mobile-friendly UI

• Shows:

  • High-risk patients
  • Priority alerts
  • Action recommendations

---

🔐 Secure & Scalable

• JWT-based authentication

---

⚡ Fast Decision Making

• Graph queries eliminate complex joins
• Enables real-time healthcare decisions

---

---

🧠 How TigerGraph Powers Sehat Dori (Technical Deep Dive)

At the heart of Sehat Dori lies a powerful idea:

💡 Healthcare is not isolated data — it is a network of relationships.

This is where TigerGraph becomes our core intelligence engine.

---

🔗 Why Graph Over Traditional Databases?

Traditional databases store data in tables:

• Patients table
• Visits table
• Hospitals table

To answer a real-world question like:

“Which pregnant women missed checkups AND have anemia AND are far from hospitals?”

You need:

• Multiple joins
• Complex queries
• Slow execution ❌

---

🚀 With TigerGraph:

We model everything as a connected graph:

(Patient) ── HAS_RISK ──> (Anemia)
     │
     ├── VISITED ──> (Visit: missed)
     │
     └── ASSIGNED_TO ──> (ASHA Worker)

👉 Now, this becomes a simple traversal problem — not a join problem.

---

🏗️ Our Graph Schema

We designed a domain-specific healthcare graph:

🔹 Vertex Types

• Patient
• Hospital
• Visit
• User (for authentication)

---

🔹 Edge Types

• Patient → HAS_RISK → RiskFactor
• Patient → VISITED → Visit
• Visit → AT → Hospital
• Patient → ASSIGNED_TO → ASHA_Worker

---

⚙️ How We Use TigerGraph (Step-by-Step)

---

1️⃣ Data Modeling

Each real-world entity is stored as a vertex, and relationships are stored as edges.

Example:

Patient (Sita)
→ HAS_RISK → Anemia  
→ VISITED → Missed Visit  
→ ASSIGNED_TO → ASHA Worker  

---

2️⃣ Query-Based Intelligence (GSQL)

Instead of raw queries, we use installed GSQL queries as APIs.

🔥 Example: High-Risk Detection

CREATE QUERY highRiskPatients() FOR GRAPH health_graph {
  Start = {Patient.*};

  HighRisk =
    SELECT p
    FROM Start:p -(HAS_RISK)-> r
    WHERE r.type == "anemia" AND p.hemoglobin < 9;

  PRINT HighRisk;
}

---

3️⃣ Multi-Factor Risk Scoring

We extend logic using accumulators:

CREATE QUERY priorityPatients() FOR GRAPH health_graph {
  Start = {Patient.*};

  Result =
    SELECT p
    FROM Start:p -(VISITED)-> v
    WHERE v.status == "missed"
    ACCUM p.@missed += 1;

  HighPriority =
    SELECT p
    FROM Result:p
    WHERE p.@missed >= 2;

  PRINT HighPriority;
}

👉 This allows real-time prioritization based on behavior patterns.

---

4️⃣ TigerGraph as an API Layer

TigerGraph automatically exposes queries as REST APIs:

GET /query/health_graph/highRiskPatients

---

5️⃣ Backend Integration

Our Node.js backend connects using HTTP:

axios.get(
  `${TG_HOST}/query/health_graph/highRiskPatients`,
  {
    headers: {
      Authorization: `Bearer ${TG_TOKEN}`
    }
  }
);

👉 No ORM, no joins — just direct graph intelligence

---

🔄 End-to-End Flow

ASHA App → Node API → TigerGraph Query  
         → Graph Traversal → Risk Detection  
         → Response → Alert 🚨

---

⚡ Performance Advantage

Traditional DB	TigerGraph
Heavy joins ❌	Native traversal ✅
Slower queries ❌	Real-time insights ✅
Rigid schema ❌	Flexible relationships ✅

---

🎯 Real Impact

With TigerGraph, MomsMagic can:

• ✅ Identify high-risk pregnancies instantly
• ✅ Track patient journeys across multiple touchpoints
• ✅ Enable proactive intervention instead of reactive care

---

🏁 Final Thought

“In healthcare, relationships matter more than records.”

TigerGraph helps us see those relationships — and act on them.

---

📊 Example Use Case

👩 Sita (7 months pregnant)

• Missed 2 checkups
• Hemoglobin = 8.5
• Lives far from hospital

👉 Sehat Dori flags her as:

🚨 HIGH PRIORITY PATIENT

👉 ASHA worker gets instant alert → takes action → life saved

---

🧪 API Example

Get High-Risk Patients

GET /api/patients/high-risk

Response:

{
  "success": true,
  "data": [
    {
      "name": "Sita",
      "risk": "HIGH"
    }
  ]
}

---

🛠️ Installation & Setup

1️⃣ Clone Repo

git clone https://github.com/your-username/momsmagic.git
cd sehat-dori

---

2️⃣ Backend Setup

cd backend
npm install

Create .env:

PORT=5000
JWT_SECRET=your_secret

TG_HOST=https://your-instance.tgcloud.io
TG_GRAPH=health_graph
TG_TOKEN=your_api_key

Run server:

npm run dev

---

3️⃣ Frontend Setup

cd frontend
npm install
npm run dev

---

🔐 Authentication Flow

• Signup → User stored in TigerGraph
• Login → JWT issued
• Protected routes → Verified using middleware

---

🧠 Why TigerGraph?

Traditional DBs:

• ❌ Struggle with multi-relationship queries
• ❌ Slow joins

TigerGraph:

• ✅ Native graph traversal
• ✅ Real-time insights
• ✅ Perfect for healthcare relationships

---

🌍 Impact

SehatDori is not just a project — it's a life-saving system.

🎯 Goals:

• Reduce maternal mortality
• Improve rural healthcare connectivity
• Empower ASHA workers with data

---

🚀 Future Scope

• 🤖 AI-based risk scoring
• 📡 Offline-first support for rural areas
• 🗺️ Geo-based emergency routing

---

👨‍💻 Team - Byte Blaze

Built with ❤️ for real-world impact.

• Sundaram Katare (Leader)
• Aryan Routela
• Khushi Sharma
• Tanmay Kanchan

---

🏁 Final Thought

“Data doesn’t save lives — connected data does.”

Sehat Dori turns connections into care.

---

⭐ If you like this project, give it a star!