parrec2dcm

Convert Philips PAR/REC files to DICOM.

Given only a .par/.rec pair (no Philips database or scanner export available), produces standards-conformant DICOM with correct geometry, UIDs, and pixel data. Two output formats are supported:

• Enhanced MR (default) — one multiframe DICOM file per series, built via highdicom's legacy conversion (SOP Class 1.2.840.10008.5.1.4.1.1.4.4, "Legacy Converted Enhanced MR Image Storage").
• Classic (--classic) — one single-frame MR Image Storage file per image, matching how most legacy PAR/REC converters behave.

UIDs are deterministic by default (the same PAR content always produces the same UIDs); pass --randomize-uids to get a fresh UID on every run instead (useful when repeatedly re-exporting the same source into the same PACS).

If a Philips "database" parameter export is also available (a flat JSON array of {Name, Value, ...} records — e.g. from a MongoDB/Agora export, or the same list gtagora-connector-py's Dataset.get_parameters() returns), pass it with --params to enrich the output beyond what the PAR header alone can provide. Currently only the "new" (REL11) database format — where the DICOM tag is embedded in the parameter name, e.g. RFR_SERIES_tag_00180091 — is mapped; the older symbolic-name format is detected and produces a warning with a graceful PAR-only fallback rather than an error. "Goal" (pulse-sequence) parameter enrichment is not yet implemented. --params accepts one or more files, which are merged into a single pool — pass a combined export or several separate ones, either way.

Installation

pip install parrec2dcm

Usage

As a command-line tool:

parrec2dcm path/to/scan.par -o path/to/output_dir              # Enhanced MR (default)
parrec2dcm path/to/scan.par -o path/to/output_dir --classic    # classic single-frame
parrec2dcm path/to/scan.par -o path/to/output_dir --params db.json   # + database enrichment

As a library:

from parrec2dcm import convert

convert("path/to/scan.par", "path/to/output_dir")
convert("path/to/scan.par", "path/to/output_dir", output_format="classic")
convert("path/to/scan.par", "path/to/output_dir", params=["db.json"])

Development

pip install -e ".[dev]"
pytest

License

MIT