qmri-io 0.1.0

NIFTI, DICOM, and BIDS I/O for qmri

qmri-io

NIFTI, DICOM, and BIDS I/O for qmri.

Installation

pip install qmri-io

Quick Start

from qmri.diffusion import adc
from qmri.io import nifti

# Load DWI data
dwi = nifti.load("dwi.nii.gz")

# Fit ADC
result = adc.fit(dwi.data, dwi.b_values, method="iwlls")

# Save result
nifti.save(result.adc, dwi.affine, "adc_map.nii.gz")

Features

• NIFTI support via nibabel
• BIDS dataset loading with automatic sidecar parsing
• DICOM support (optional, requires pydicom)
• Sidecar file handling (.bval, .bvec, .json)

Design

This package handles file I/O so the core qmri package doesn't need nibabel or pydicom as dependencies. Functions return simple dataclasses with numpy arrays, not complex container objects.

Documentation

Full documentation: https://gold-standard-phantoms.github.io/qmri

Licence

MIT