Skip to content

Preclinical MRI Dataset Toolkit

Documentation for BrkRaw.

BrkRaw is a toolkit for loading Bruker Paravision MRI studies, inspecting and normalizing metadata, mapping parameters through rule and spec systems, and exporting data to NIfTI with optional sidecar metadata.

It provides both a CLI and a Python API, and is designed to be extended through add-on rules, specs, and converter plugins.

Highlights

  • Load Bruker Paravision datasets from directories, zip archives, or Paravision-exported .PvDatasets files
  • Inspect study, scan, and reconstruction metadata via rich CLI tables
  • Map parameters using rule-based selection and remapper specs
  • Convert scans to NIfTI with configurable layout entries and templates
  • Extend conversion behavior via rules, specs, and converter hooks

Supported scope and limitations

BrkRaw focuses on preclinical MRI datasets acquired with Bruker Paravision. The current scope is intentionally constrained to ensure correctness and reproducibility.

Supported

  • Bruker Paravision datasets (PV5, PV6, PV7, PV360)
  • Paravision-exported .PvDatasets archives
  • Preclinical MRI acquisitions (2D and 3D)
  • Multiple reconstructions (reco_id) per scan
  • Multi-slicepack acquisitions (exported as separate outputs with suffixes)
  • Loading from directory trees, zip archives, or .PvDatasets files
  • Export to NIfTI with optional JSON sidecar metadata

Tested Paravision versions

BrkRaw has been validated using Bruker Paravision standard datasets exported from the following versions:

  • Paravision PV 360.3.5 - 360.3.7

Standard dataset definitions are based on Bruker documentation: https://www.bruker.com/en/products-and-solutions/preclinical-mri/paravision

Support for other Paravision versions may work but has not been systematically validated.

Not in scope / limitations

  • Clinical DICOM datasets
  • Automatic DICOM-to-NIfTI conversion
  • Vendor-agnostic MRI pipelines
  • Full BIDS compliance without user-provided rules or specs
  • Interactive visualization (handled by external viewers)

BrkRaw intentionally separates data interpretation from conversion rules. Users are expected to encode project- or modality-specific logic through rules, specs, or extensions rather than hard-coded heuristics.

Ecosystem and extensions

BrkRaw is designed as a core engine with a growing ecosystem of extensions and supporting tools.

Official resources

Converter and UI extensions

Extension templates