mdreg.fetch#

mdreg.fetch(dataset=None, clear_cache=False, download_all=False) → dict[source]#

Fetch a dataset included in mdreg

Parameters:

  • dataset (str, optional) – name of the dataset. See below for options.

  • clear_cache (bool, optional) – When a dataset is fetched, it is downloaded and then stored in a local cache memory for faster access next time it is fetched. Set clear_cache=True to delete all data in the cache memory. Default is False.

  • download_all (bool, optional) – By default only the dataset that is fetched is downloaded. Set download_all=True to download all datasets at once. This will cost some time but then offers fast and offline access to all datasets afterwards. This will take up around 300 MB of space on your hard drive. Default is False.

Returns:

Data as a dictionary.

Return type:

dict

Example

Fetch the MOLLI images, and display as animation:

>>> import mdreg
>>> import mdreg.plot as plt

Get the data:

>>> data = mdreg.fetch('MOLLI')

Plot as animation:

>>> plt.animation(data['array'], vmin=0, vmax=1e4)

(Source code, png, hires.png, pdf)

../../_images/mdreg-fetch-1.png

Notes

The following datasets can be fetched:

MOLLI

Size: 2MB

Background: T1-mapping data for the kidney acquired on a healthy volunteer, collected as part of the technical validation efforts of the UKRIN-MAPS consortium and the development of the UKAT package.

Data format: The fetch function returns a dictionary, which contains the following items:

  • array: 4D array of signal intensities in the abdomen at different inversion times.

  • TI: A list of inversion times in msec.

Funding statement:

Data collection was funded by the UKRIN-MAPS MRC Partnership grant (MR/R02264X/1) and the NIHR AFiRM project (NIHR128494).

MOLLI_small

Size: 129KB

Background: A small version of ‘MOLLI’ dataset, in 4x lower resolution than the original. Useful for rapid testing and debugging on a local machine.

MOLLI_tiny

Size: 9KB

Background: A tiny version of ‘MOLLI’ dataset, in 16x lower resolution than the original. Useful for testing and debugging solutions on remote machines.

VFA

Size: 5MB

Background: 4D variable flip angle data for T1-mapping in the abdomen. Data are provided by the liver work package of the TRISTAN project which develops imaging biomarkers for drug safety assessment. The data and analysis was first presented at the ISMRM in 2024 (Min et al 2024, manuscript in press). A single set of variable flip angle data are included.

Data format: The fetch function returns a dictionary, which contains the following items:

  • array: 4D array of signal intensities in the liver at different flip angles

  • FA: flip angles in degrees

  • spacing: voxel size in mm in x-, y-, and z-directions as an array.

Please reference the following abstract when using these data:

Thazin Min, Marta Tibiletti, Paul Hockings, Aleksandra Galetin, Ebony Gunwhy, Gerry Kenna, Nicola Melillo, Geoff JM Parker, Gunnar Schuetz, Daniel Scotcher, John Waterton, Ian Rowe, and Steven Sourbron. Measurement of liver function with dynamic gadoxetate-enhanced MRI: a validation study in healthy volunteers. Proc Intl Soc Mag Reson Med, Singapore 2024.

VFA_small

Size: 87KB

Background: A small version of ‘VFA’ dataset, in 5x lower resolution than the original. Useful for rapid testing and debugging on a local machine.

VFA_tiny

Size: 7KB

Background: A tiny version of ‘VFA’ dataset, in 10x lower resolution than the original. Useful for testing and debugging cloud-based solutions that consume funds in proportion to data size.

On this page

This Page