Follow these instructions to batch download the Stanford 2D FSE and Stanford Fullysampled 3D FSE Knees datasets. These datasets can be found here, however batch download is currently not supported on the website.
Here I describe how to download the datasets using the mridata
Python command line tool. The MRAugment repository uses the fastMRI codebase to handle MRI datasets, therefore we need to convert the downloaded datasets to have a format similar to fastMRI data. This is done through tools from mridata-recon repository.
The size of the datasets is shown in the following table.
Original | Converted | |
---|---|---|
Stanford 2D FSE | 40G | 37G |
Stanford 3D FSE | 32G | 34G |
- Python3
-
In order to install
mridata
first you need to installrequests
,tqdm
, andboto
usingpip install requests boto3 tqdm
-
Next, install
mridata
usingpip install mridata
-
Download the dataset. Navigate to the directory where you want to download the dataset and make sure that the correct UUID text file (Stanford_2D_TSE_uuid.txt or Stanford_3D_FSE_knee_uuid.txt) is saved in the same folder. You can find both of these text files in this repository. To dowload the datasets run the command
mridata batch_download Stanford_2D_FSE_uuid.txt
or
mridata batch_download Stanford_3D_FSE_knee_uuid.txt
depending on which dataset you would like to download.
-
Convert the datasets using the scripts in this repository. Run
python convert_stanford2d.py --input_dir --output_dir
or
python convert_stanford3d.py --input_dir --output_dir
and replace the input and output directories with the desired path on your machine.
Both of these datasets can be downloaded directly from http://mridata.org/
- Stanford 2D FSE: Joseph Y. Cheng, https://github.com/MRSRL/mridata-recon/
- Stanford Fullysampled 3D FSE Knees: Epperson K, Sawyer AM, Lustig M, Alley M, Uecker M., Creation Of Fully Sampled MR Data Repository For Compressed Sensing Of The Knee. In: Proceedings of the 22nd Annual Meeting for Section for Magnetic Resonance Technologists, 2013
mridata
command line tool repository: https://github.com/mikgroup/mridata-python- Tools used to extract kspace data from original datasets: https://github.com/MRSRL/mridata-recon/