EmergencyX is an intelligent desktop application that leverages Machine Learning to revolutionize patient triage in emergency departments — predicting emergency severity levels and cardiac arrest risk in real time.
- Overview
- Key Features
- System Architecture
- Modules
- Tech Stack
- Installation
- How to Use
- ML Model Details
- Project Structure
- Screenshots
- Contributing
- Author
EmergencyX addresses one of the most critical challenges in emergency medicine — speed and accuracy of triage. Every second counts in an emergency, and misclassifying a patient's severity can be fatal.
By combining a modern Windows desktop UI with a trained Machine Learning backend, EmergencyX empowers medical staff to:
- Instantly assess emergency severity for incoming patients
- Predict cardiac arrest risk based on vital signs and clinical data
- Prioritize patients with confidence backed by data-driven AI
This project demonstrates the real-world integration of C# desktop development with ML-based predictive modeling.
| Feature | Description |
|---|---|
| 🏥 Smart Triage | Predicts emergency level (1–5 scale) based on patient vitals & symptoms |
| ❤️ Cardiac Risk Prediction | Detects risk of cardiac crisis using clinical indicators |
| 🤖 ML Integration | C# frontend communicates with trained ML model for real-time inference |
| 🖥️ Desktop Application | Native Windows app — fast, offline-capable, no browser needed |
| 📊 Instant Results | Predictions displayed within milliseconds of data entry |
| 🎯 Dual Module System | Two independent prediction modules in one unified interface |
┌─────────────────────────────────────────────────────────────┐
│ EmergencyX Desktop App │
│ (C# / .NET) │
│ │
│ ┌─────────────────────┐ ┌─────────────────────────┐ │
│ │ Module 1 │ │ Module 2 │ │
│ │ Emergency Triage │ │ Cardiac Crisis │ │
│ │ (Patient Input) │ │ Prediction │ │
│ └────────┬────────────┘ └────────────┬────────────┘ │
│ │ │ │
│ └──────────┬──────────────────┘ │
│ │ │
│ ┌────────▼────────┐ │
│ │ ML Bridge │ │
│ │ (API / .NET) │ │
│ └────────┬────────┘ │
└───────────────────────┼─────────────────────────────────────┘
│
┌─────────────▼──────────────┐
│ ML Model Backend │
│ (ML.NET / Python / ONNX) │
│ │
│ ┌──────────┐ ┌──────────┐ │
│ │ Triage │ │ Cardiac │ │
│ │ Model │ │ Model │ │
│ └──────────┘ └──────────┘ │
└────────────────────────────┘
The triage module accepts patient data and classifies the emergency level using a trained classification model.
Input Parameters:
- Age, Gender
- Chief complaint / Symptoms
- Vital signs (Heart Rate, Blood Pressure, O₂ Saturation, Temperature, Respiratory Rate)
- Consciousness level (GCS)
- Pain score (0–10)
Output:
- Emergency Level:
Level 1 (Resuscitation)→Level 5 (Non-Urgent) - Confidence score (%)
- Recommended action
The cardiac module analyzes clinical indicators to assess the risk of an acute cardiac event.
Input Parameters:
- Age, Gender, BMI
- Chest pain type & duration
- ECG findings
- Cardiac biomarkers (Troponin, CK-MB)
- Blood pressure, Cholesterol
- History of diabetes, hypertension, smoking
Output:
- Risk Level:
Low/Moderate/High/Critical - Probability score (%)
- Alert flag for immediate intervention
Frontend (Desktop UI)
├── Language : C# (.NET 6 / .NET Framework)
├── UI Framework : Windows Forms / WPF
└── IDE : Visual Studio 2022
Machine Learning Backend
├── ML Framework : ML.NET ──or── Python (scikit-learn / XGBoost)
├── Model Format : .ZIP (ML.NET) ──or── .pkl / .onnx (Python)
├── Integration : REST API (Flask) ──or── Direct ML.NET call
└── Training Data : Emergency & Cardiac clinical datasets
Communication Layer
├── If Python ML : HTTP REST API (localhost Flask server)
└── If ML.NET : Direct in-process model loading
- Windows 10 / 11 (64-bit)
- .NET 6.0 Runtime or higher
- Python 3.8+ (only if using Python ML backend)
- Visual Studio 2022 (for development)
git clone https://github.com/YOUR_USERNAME/EmergencyX.git
cd EmergencyX# Open solution in Visual Studio
start EmergencyX.sln
# Or build via CLI
dotnet restore
dotnet build
dotnet runcd MLBackend
pip install -r requirements.txt
python app.pyThe Flask server will start on
http://localhost:5000and the C# app will connect automatically.
1. Launch the Application
Run
EmergencyX.exefrom thebin/Releasefolder or pressF5in Visual Studio.
2. Select a Module
Choose between
Emergency TriageorCardiac Crisis Predictionfrom the main dashboard.
3. Enter Patient Data
Fill in the patient's clinical information in the structured form.
4. Run Prediction
Click "Analyze" — the ML model processes the input and returns results instantly.
5. Review Results
View the predicted severity level, confidence score, and recommended clinical action.
| Property | Triage Model | Cardiac Model |
|---|---|---|
| Type | Multi-class Classification | Binary / Multi-class Classification |
| Algorithm | Random Forest / XGBoost | Logistic Regression / Neural Net |
| Output Classes | 5 Emergency Levels | 4 Risk Levels |
| Input Features | ~10 clinical features | ~12 clinical features |
| Model Format | .zip (ML.NET) / .pkl |
.zip (ML.NET) / .pkl |
EmergencyX/
│
├── EmergencyX/ # Main C# Application
│ ├── Forms/
│ │ ├── MainDashboard.cs # Main window
│ │ ├── TriageForm.cs # Module 1 UI
│ │ └── CardiacForm.cs # Module 2 UI
│ ├── Models/
│ │ ├── PatientData.cs # Patient data model
│ │ └── PredictionResult.cs # Result model
│ ├── Services/
│ │ ├── MLService.cs # ML model communication
│ │ └── ApiService.cs # REST API client (if Python)
│ └── EmergencyX.csproj
│
├── MLBackend/ # Python ML Backend (optional)
│ ├── models/
│ │ ├── triage_model.pkl
│ │ └── cardiac_model.pkl
│ ├── app.py # Flask API server
│ ├── train_triage.py # Model training script
│ ├── train_cardiac.py # Model training script
│ └── requirements.txt
│
├── MLModels/ # ML.NET Models (alternative)
│ ├── TriageModel.zip
│ └── CardiacModel.zip
│
├── Screenshots/ # App screenshots
├── README.md
├── .gitignore
└── LICENSE
| Dashboard | Triage Module | Cardiac Module |
|---|---|---|
 |
 |
 |
# Visual Studio
.vs/
bin/
obj/
*.user
*.suo
# Python
__pycache__/
*.pyc
*.pyo
.env
venv/
*.pkl # optional: ignore trained models if large
# ML.NET models (if large, use Git LFS instead)
# *.zip
# OS
.DS_Store
Thumbs.db
*.logContributions, issues and feature requests are welcome!
- Fork the repository
- Create your feature branch:
git checkout -b feature/AmazingFeature - Commit your changes:
git commit -m 'Add AmazingFeature' - Push to the branch:
git push origin feature/AmazingFeature - Open a Pull Request
Ahmad Albasha
This project is licensed under the MIT License — see the LICENSE file for details.
⭐ If you find this project useful, please consider giving it a star!
Built with ❤️ to save lives through technology