spchetan1987/hospitalqueuemanager

🏥 AI Hospital Queue Management System (ML-Powered)

An intelligent hospital queue management system that uses Machine Learning to prioritize patients based on age and critical symptoms.

🎯 Features

• —ML-Based Prioritization: Uses a Random Forest Classifier trained on 5000 samples
• —Real-time Queue Management: Automatically sorts patients by priority
• —Confidence Scores: Shows prediction confidence for transparency
• —Interactive UI: Built with Gradio for easy use
• —Feature Engineering: Extracts meaningful features from age and symptoms

🧠 ML Model Details

Model Performance

• —Training Accuracy: 92.37%
• —Test Accuracy: 91.10%
• —Algorithm: Random Forest Classifier (100 trees)
• —Training Samples: 5000 synthetic patient records

Features Used

1. 1.Age: Patient's age
2. 2.Has Critical Symptom: Binary indicator for life-threatening symptoms
3. 3.Has Medium Symptom: Binary indicator for moderate symptoms
4. 4.Is Elderly: Age > 65
5. 5.Is Child: Age < 12
6. 6.Is Infant: Age < 5
7. 7.Age Normalized: Age scaled to 0-1 range
8. 8.Symptom Word Count: Complexity of symptom description

Priority Levels

• —Priority 1 (HIGH): Critical symptoms requiring immediate attention
• —Examples: chest pain, shortness of breath, severe bleeding, unconscious, heart attack, stroke
• —Priority 2 (MEDIUM): Moderate symptoms or vulnerable age groups
• —Examples: high fever, fracture, vomiting, elderly patients, infants
• —Priority 3 (LOW): Minor symptoms that can wait
• —Examples: cold, cough, minor headache, rash, routine check-up

📦 Installation

Prerequisites

• —Python 3.7 or higher
• —pip package manager

Steps

1. 1.Clone or download this repository

1. 1.Install dependencies:

pip install -r requirements.txt

1. 1.Train the ML model (if not already trained):

python train_model.py

This will:

• —Generate 5000 synthetic training samples
• —Train a Random Forest model
• —Save the model to queue_priority_model.pkl
• —Save configuration to model_config.json
• —Create training_data.csv with the training dataset

1. 1.Run the application:

python app.py

The Gradio interface will launch in your default browser (usually at http://127.0.0.1:7860)

🚀 Usage

Adding Patients to Queue

1. 1.Enter the patient's Name
2. 2.Enter the patient's Age (0-120)
3. 3.Enter Symptoms (e.g., "chest pain", "high fever", "minor headache")
4. 4.Click "Add to Queue"

The ML model will:

• —Extract features from the input
• —Predict the priority level (1, 2, or 3)
• —Show confidence scores for transparency
• —Add the patient to the queue sorted by priority

Example Inputs

High Priority Patient:

• —Name: John Doe
• —Age: 55
• —Symptoms: chest pain and shortness of breath

Medium Priority Patient:

• —Name: Mary Smith
• —Age: 72
• —Symptoms: high fever and dizziness

Low Priority Patient:

• —Name: Tom Brown
• —Age: 25
• —Symptoms: minor headache and cough

Clearing the Queue

Click the "Clear Queue" button to reset the patient queue.

📊 Model Training Details

Data Generation

The training data is synthetically generated with realistic distributions:

• —25% High Priority cases
• —25% Medium Priority cases
• —50% Low Priority cases

Special rules applied:

• —Elderly patients (>65) with non-critical symptoms → Medium priority
• —Infants (<5) with any symptoms → At least Medium priority
• —Critical symptoms → Always High priority regardless of age

Feature Importance

Based on the trained model:

1. 1.Has Critical Symptom: 53.96% importance
2. 2.Has Medium Symptom: 34.94% importance
3. 3.Age: 3.24% importance
4. 4.Symptom Word Count: 3.23% importance
5. 5.Other features: < 3% each

🔧 Project Structure

hospital-queue-management/
├── app.py                      # Main Gradio application
├── train_model.py              # ML model training script
├── requirements.txt            # Python dependencies
├── queue_priority_model.pkl    # Trained ML model (generated)
├── model_config.json          # Model configuration (generated)
├── training_data.csv          # Training dataset (generated)
└── README.md                  # This file

🛠️ Technical Stack

• —Frontend: Gradio 5.49.1
• —ML Framework: scikit-learn
• —Data Processing: pandas, numpy
• —Model Persistence: joblib

📈 Future Enhancements

• —Add more features (vital signs, medical history)
• —Implement real-time model retraining
• —Add patient wait time estimation
• —Include doctor availability in prioritization
• —Export queue data to CSV/Excel
• —Multi-language support
• —Integration with hospital management systems

🤝 Contributing

Feel free to fork this project and submit pull requests for improvements!

📄 License

This project is open source and available for educational purposes.

🙏 Acknowledgments

Built with Gradio and deployed on Hugging Face Spaces

Note: This is a demonstration system. For production use in real hospitals, please consult with medical professionals and ensure compliance with healthcare regulations (HIPAA, etc.).