🏥 HealthCheckAI — Predictive Patient Risk Monitoring

HealthCheckAI is an end-to-end machine learning project that predicts the risk of ICU transfer for hospital patients using vital signs and lab results.

---

📦 Features

• Predicts patient deterioration risk using a Random Forest Classifier
• Trained on synthetic EHR data (age, vitals, labs, etc.)
• SHAP explainability for feature attribution
• Streamlit frontend for live clinical risk prediction
• REST API via FastAPI for system integration
• Docker & Docker Compose setup for deployment

---

🚀 Quickstart

1. Clone and Build

git clone https://github.com/yourusername/healthcheckai.git
cd healthcheckai
docker-compose up --build

• Streamlit: http://localhost:8501
• FastAPI Docs: http://localhost:8000/docs

---

📊 Dataset

The project uses synthetic data with features like:

• age, gender, admission_type, comorbidities
• heart_rate, respiratory_rate, systolic_bp, temperature
• hemoglobin, creatinine, lactate, white_cell_count

Label: risk_label = 1 indicates ICU transfer risk.

---

🛠 Tech Stack

• Python, Pandas, scikit-learn, SHAP
• Streamlit, FastAPI, Docker
• Synthetic data generated for reproducibility

---

📂 File Structure

├── data/                  # Raw and processed datasets
├── models/                # Trained model, scaler, SHAP plots
├── app/
│   ├── streamlit_app.py   # Streamlit UI
│   └── api.py             # FastAPI backend
├── notebooks/             # EDA and feature engineering
├── Dockerfile             # Container build
├── docker-compose.yml     # Multi-service orchestration
├── requirements.txt       # Python dependencies
└── README.md              # Project overview

---

👩‍⚕️ Example Prediction

Send a POST request to /predict:

{
  "age": 70,
  "comorbidities": 3,
  "heart_rate": 105,
  "respiratory_rate": 25,
  "systolic_bp": 85,
  "temperature": 38.1,
  "hemoglobin": 11.5,
  "creatinine": 1.2,
  "lactate": 3.0,
  "white_cell_count": 9.5,
  "gender": "Male",
  "admission_type": "Emergency"
}

---

📜 License

MIT