Description of the data and file structure

This repository contains raw MRI data comparing strategies for achieving widespread delivery of small molecules across the blood–brain barrier (BBB) in rats: chemically mediated BBB disruption (LPA), ultrasound-mediated BBB disruption, and cerebrospinal fluid (CSF) infusion via the cisterna magna, alongside a gadoteridol-only control delivered via carotid catheter infusion. Data capture the spatial distribution and temporal dynamics of the contrast agent gadoteridol across brain regions, supporting quantitative analysis of delivery efficiency, regional accessibility, and transport kinetics for each method.

Data are organized into one subfolder per animal within mri_datasets.zip, grouped by treatment condition (see below). Each animal folder contains one or more scan_## subfolders, each holding that scan's 2dseqfile.

Software requirements and data access

Important: The MRI data in this repository are stored in a proprietary binary format (Bruker ParaVision) and cannot be opened directly. Each scan folder contains data in a 2dseq (reconstructed image data) format. Use one of the methods below to access them.

Recommended: MATLAB with the provided loadbruker2.m script

The recommended way to access these data is using MATLAB with the loadbruker2.m script provided in this repository. MATLAB is a commercial product; a free 30-day trial is available at mathworks.com.

General usage:

A = loadbruker2(filename, [x y z t])

• filename: path to the 2dseq file (e.g., 'us_02/scan_04/2dseq')
• x, y: image dimensions in pixels (x=75, y=150 for this dataset)
• z: number of slices (or, here, echo images) — 5
• t: number of time points (or, here, TR values) — 8

For this dataset, every scan shares the same fixed acquisition parameters (see Scan parameters below).

Alternative: Python (open source)

Users without MATLAB access can use the open-source brukerapi Python library, which reads 2dseqfiles directly. Documentation is available at bruker-api.readthedocs.io. The open-source bruker2nifti and BrkRaw packages are also available if converting to NIfTI is preferred.

Alternative: Bruker ParaVision

The proprietary acquisition software used to generate these files. Institutional MRI core facilities may have licensed access.

Bruker directory structure

Each animal folder (e.g., us_02) contains one or more scan_## subfolders, one per scan. Within each scan_## folder is the following file:

• 2dseq — Binary file containing the reconstructed image data. Load using the loadbruker2.m script (see above) or the Python brukerapi library, using the fixed dimensions given in Scan parameters below.

Files and variables

File: mri_datasets.zip

Each rat is organized by treatment condition:

1. cm — Intra-CSF delivery of gadoteridol via cisterna magna injection.
2. gd — Injection of gadoteridol via carotid catheter infusion.
3. lpa — Co-injection of lysophosphatidic acid (LPA) and gadoteridol via carotid catheter infusion.
4. us — Single ultrasound sonication followed by gadoteridol via tail vein infusion.
5. mus — Multiple ultrasound sonications followed by gadoteridol via tail vein infusion.

Each animal folder (e.g., us_02) contains one or more scan_## subfolders, each with that scan's 2dseq file (loadable via loadbruker2.m, see above).

Scan parameters (all conditions and all scans):

• Scan protocol: RAREVTR on a Bruker BioSpec MRI
• Matrix: 75x150 (FOV: 25x50 mm)
• TE = 10, 30, 50, 70, 90 ms (5 values)
• TR = 5000, 3000, 1500, 1000, 500, 250, 150, 120 ms (8 values)
• RARE factor: 2

Loading example (MATLAB):

A = loadbruker2('us_02/scan_04/2dseq', [75 150 5 8]);

Access information

Other publicly accessible locations of the data: N/A

Data was derived from the following sources: N/A