ormir-mids

I/O of Medical Image Data Structure (MIDS) for Open and Reproducible Musculoskeletal Imaging Research (ORMIR). Based on the BIDS data structure.

Note

This package is a fork of muscle-bids for muscle MR imaging data.

Tutorial

Run the tutorial:

Main contributors

• Francesco Santini
• Donnie Cameron
• Leonardo Barzaghi
• Judith Cueto Fernandez
• Jilmen Quintiens
• Lee Youngjun
• Jukka Hirvasniemi
• Gianluca Iori

Installation

Dependencies

To install ormir-mids, run the code below, noting this list of dependencies.

It is recommended to install ormir-mids in a separate virtual environment:

conda env create -n ormir-mids
conda activate ormir-mids

Clone the git repository:

git clone https://github.com/ormir-mids/ormir-mids.git

Now we can install the package using pip. This will also install the required dependencies.

cd ormir-mids
pip install .
pip install --upgrade nibabel # the default nibabel has bugs

Usage

ormir-mids can be used in two ways:

1. Running dcm2mbids as an executable to convert DICOM data to the BIDS format.
2. Importing ormir-mids as a Python module to find, load, and interrogate BIDS-format data.

The commandline script is called dcm2mbids.exe. To view the commandline script help type

dcm2mbids -h

To use ormir-mids within Python, import the following modules

from muscle_bids.utils.io import find_bids, load_bids
import nibabel as nib

For a detailed description of how to use ormir-mids see the following notebook

ormir-mids usage: dcm2mbids

Medical volume manipulation within Pyhon

ormir-mids can be used within Python to load, manipule and visualize medical volume datasets, without convertion to the BIDS format.

• Load a DICOM file to a MedicalVolume object

from muscle_bids.utils.io import load_dicom

mv = load_dicom('<Path-to-DICOM-file>')

• Slice the volume. This will create a separate subvolume. Metadata will be sliced appropriately.

mv_subvolume = mv[50:90, 50:90, 30:70]

• Convert volume to ITK image with SimpleITK

mv_itk = mv.to_sitk()

Examples of how to use ormir-mids for common data handling, image manipulation and processing tasks within Python can be found in this notebook

ormir-mids usage: MedicalVolume class

Acknowledgement

This package was developed thanks to the support of the JCMSK community during the Maastricht 2022 workshop and hackaton “Building the Jupyter Community in Musculoskeletal Imaging Research”.

Image I/O is based on DOSMA by Arjun Desai. A stripped-down version of DOSMA is present in muscle-bids/dosma_io.