CDSS App

Flutter + FastAPI CT Analysis System

---

📌 Overview

CDSS App is a Flutter-based medical image analysis application that connects to a locally hosted FastAPI backend.

The app:

• Sends multiple CT scan images to a FastAPI server
• Receives a structured medical report (text format)
• Converts the report into an image
• Displays and stores the result locally

---

🔄 Complete Workflow

1. User selects multiple images
2. App sends images to FastAPI backend (/analyze)
3. Backend processes images using AI model
4. Backend returns JSON response containing:
   • success
   • refined_report
5. App converts the report text into a PNG image
6. Result is displayed and saved in local database

---

🏗️ System Architecture

Flutter App
     ↓
POST /analyze (multipart/form-data)
     ↓
FastAPI Server (Laptop)
     ↓
AI Model Processing
     ↓
JSON Response (refined_report)
     ↓
Flutter converts text → Image
     ↓
Display + Store in DB

---

⚙️ Requirements

• Python 3.8+
• FastAPI
• Uvicorn
• python-multipart
• Flutter SDK
• Laptop and phone connected to same WiFi network

---

🚀 Backend Setup

1️⃣ Install Dependencies

pip install fastapi uvicorn python-multipart

---

2️⃣ Run Server

uvicorn app:app --host 0.0.0.0 --port 8000

0.0.0.0 is required to allow phone access.

---

3️⃣ Get Laptop IPv4 Address

Open Command Prompt:

ipconfig

Find:

IPv4 Address : 10.115.206.17

Your API base URL becomes:

http://10.115.206.17:8000

---

🌐 API Endpoint Used in App

POST /analyze

POST http://<YOUR_IP>:8000/analyze

Request Type

• Method: POST
• Content-Type: multipart/form-data
• Field name: files
• Multiple images allowed

---

📥 Expected Backend Response Format

The backend must return JSON in this format:

{
  "success": true,
  "refined_report": "Detailed CT analysis report text here..."
}

If "success" is not true, the app throws an error.

---

📱 Flutter Implementation (submit_page.dart)

The app sends images using:

final uri = Uri.parse("http://10.115.206.17:8000/analyze");
final request = http.MultipartRequest("POST", uri);

for (final img in widget.images) {
  request.files.add(
    await http.MultipartFile.fromPath("files", img.path),
  );
}

After receiving response:

• JSON is decoded
• refined_report is extracted
• Text is converted into PNG image
• Image is stored in local database
• Output page is shown

---

🖼️ Text to Image Conversion

The app dynamically converts the medical report into a PNG file using Flutter Canvas API.

This ensures:

• Structured medical report view
• Saveable image output
• Easy storage in DB

---

🧪 Testing Backend

Open in browser:

http://<YOUR_IP>:8000/docs

Use Swagger UI to test /analyze endpoint.

---

🌐 Network Requirements

• Laptop and phone must be on same WiFi
• Firewall must allow port 8000
• Backend must be running
• IPv4 address must match the one in submit_page.dart

---

❗ Common Errors

Phone cannot connect

• Check same WiFi
• Verify IPv4 address
• Ensure --host 0.0.0.0
• Check firewall settings

JSON decode error

• Ensure backend returns proper JSON
• Ensure success field exists

---

📂 Project Structure

cdss_app/
│
├── lib/
│   ├── submit_page.dart
│   ├── data/
│   └── ml/
│
├── backend/
│   └── app.py
│
└── README.md

---

📌 Notes

• IP address may change when WiFi reconnects.
• For production deployment:
  • Use cloud hosting
  • Use HTTPS
  • Do not hardcode IP address

---