ajitonelson/synthetic-mass-health-data

SyntheticMass Health Data for ArangoDB Hackathon

🏆 Hackathon Context: Building the Next-Gen Agentic App with GraphRAG & NVIDIA cuGraph

This dataset has been specially processed for the ArangoDB Hackathon focused on building the future of Generative AI applications. The hackathon challenges participants to harness the power of graph technologies to create intelligent, adaptive systems using ArangoDB.

📊 Dataset Overview

This dataset contains processed CSV files from the SyntheticMass project, which provides synthetic patient medical records. The data has been extracted and transformed specifically for use with graph databases and AI applications.

📁 Files Included:

• patients.csv - Core patient demographic information
• conditions.csv - Patient health conditions and diagnoses
• encounters.csv - Healthcare provider visits and encounters
• observations.csv - Clinical observations and measurements
• medications.csv - Prescribed medications and treatments
• immunizations.csv - Vaccination records
• procedures.csv - Medical procedures performed
• allergies.csv - Patient allergy information
• careplans.csv - Patient care planning documents

🔄 Data Processing

The original data has been processed through a series of steps:

1. Extraction from SyntheticMass FHIR 3.0.1 archives
2. Conversion to CSV format
3. Merging of multiple output files
4. Optimization for graph database ingestion

The complete data processing notebook can be found here: H_ArangoDB_Download_Extract_and_Merge_Data.ipynb

📚 Data Source

This dataset is derived from the SyntheticMass project:

• Source: SyntheticMass Downloads
• Version: SyntheticMass Data, Version 2 (24 May, 2017)
• Original Format: FHIR 3.0.1, CSV, C-CDA
• Original Size: 21GB

🔗 Use Cases

This dataset is particularly suitable for:

• Building healthcare knowledge graphs
• Developing RAG (Retrieval-Augmented Generation) applications
• Training ML models for patient care prediction
• Creating intelligent healthcare assistance systems
• Demonstrating graph-based data relationships
• Prototyping medical decision support systems

📝 Citation

If you use this dataset, please cite:

• The original SyntheticMass project: synthea.mitre.org
• This processed dataset version for the ArangoDB Hackathon

🛠 Getting Started

To use this dataset with ArangoDB:

1. Download the ZIP file
2. Extract the CSV files
3. Import into ArangoDB using the appropriate collection structure
4. Ensure proper relationship mapping between collections

⚖️ License

This dataset maintains the original SyntheticMass licensing terms while being provided for hackathon use. Please refer to SyntheticMass licensing for complete terms.

🤝 Contributing

For questions or contributions regarding this processed dataset, please visit the GitHub repository.

---

Dataset prepared by ajitonelson for the ArangoDB Hackathon