fhir2meds 0.1.0

Convert HL7 FHIR (v4/v5) Observation resources to MEDS sharded Parquet format.

MEDS on FHIR

Overview

MEDS on FHIR is a Python package for converting HL7 FHIR (v4/v5) resources into the MEDS sharded Parquet format. It supports scalable, schema-validated event extraction and is designed for research and interoperability projects. This project is in its early stages and has been tested mainly with the MIMIC-IV FHIR demo.

Features

• Parses and processes all MEDS-compatible FHIR resource types (v4/v5) (tested with MIMIC-IV FHIR demo)
• Robust mapping from FHIR Observation to MEDS event schema
• Handles patient ID resolution and vocabulary mapping
• Outputs sharded Parquet files, validated against the MEDS schema
• Extensible: add mapping for new FHIR resource types easily
• Comprehensive test suite for FHIR resource parsing

---

Installation

pip install fhir2meds
# or for local development
pip install -e .

---

Usage

fhir2meds raw_input_dir=mimic-fhir root_output_dir=example_output ++overwrite=true ++verbose=true

• raw_input_dir: Directory containing FHIR .ndjson/.json files (e.g., MIMIC-IV FHIR demo)
• output_dir: Output directory for MEDS Parquet shards
• max_observations: (Optional) Limit number of observations for debugging
• overwrite: (Optional) Overwrite existing output directory
• verbose: (Optional) Enable verbose logging
• do_download: (Optional) Download MIMIC-IV FHIR demo dataset automatically (to be tested)

---

Testing

Run the test suite (requires pytest):

PYTHONPATH=src pytest tests/

---

References

• HL7 FHIR Resource List
• MIMIC-IV FHIR Demo
• MEDS Format & Schema
• ETL_MEDS_Template R4 Resource List: https://hl7.org/fhir/R4/resourcelist.html

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contributing

Pull requests are welcome! For major changes, please open an issue first to discuss what you would like to change.