BigBrain-MR: a new digital phantom with anatomically-realistic magnetic resonance properties at 100-µm resolution

BigBrain-MR is a novel digital phantom with realistic anatomical detail up to 100-µm resolution, including multiple MRI contrasts and properties that affect image generation. This phantom was generated from the publicly available BigBrain histological dataset and from lower-resolution in-vivo 7T-MRI data, using a new image processing framework that allows mapping the general properties of in-vivo data into the fine anatomical scale of BigBrain.

The dataset includes:

BigBrain original contrast and a new atlas with 20 ROIs;

T1-weighted image and T1 map;

T2*-weighted images and R2* map;

Magnetic susceptibility map (QSM);

Background magnetic field map;

Complex coil sensitivity maps (32ch-receive RF array);

Bias field map.

Information about each image/map (including data type and amplitude scaling) is provided in data_info.txt.

Additionally, we have included a script with usage examples in Python that illustrate how the data can be loaded, processed and combined for diverse simulation purposes.

BigBrain-MR is presented, described and tested in the following peer-reviewed article:

C. Sainz Martinez, M. Bach Cuadra, J. Jorge. BigBrain-MR: a new digital phantom with anatomically-realistic magnetic resonance properties at 100-µm resolution for magnetic resonance methods development. NeuroImage 2023. DOI: 10.1016/j.neuroimage.2023.120074