Skip to content

Latest commit

 

History

History
47 lines (31 loc) · 2.26 KB

README.md

File metadata and controls

47 lines (31 loc) · 2.26 KB
Raw

Preparing datasets

This README file provides instructions for using the Python script to preprocess the fastMRI dataset. The script processes HDF5 files, extracts k-space data, performs the Fourier transform to reconstruct images, and saves the images in cfl format.

Usage

Command-Line Arguments

  • --data_folder: Path to the folder containing the input .h5 files. Default is /scratch/gluo/compressed/fastMRI/multicoil_train.
  • --save_folder: Path to the folder where the processed images will be saved. Default is /scratch/gluo/fastMRI.
  • --start_id: Starting ID for naming the output files. Default is 1000000.

Running the Script

  1. Open a terminal or command prompt.
  2. Navigate to the directory containing the script.
  3. Download the BART binary file from the provided link, make it executable and set the BART_PATH environment variable to it.
  4. Run the script with the desired arguments.

For example:

wget -nv https://huggingface.co/Guanxiong/MRI-Image-Priors/resolve/main/Data/bart
chmod +x bart
export BART_PATH=$(pwd)/bart
python fastmri.py --data_folder /path/to/input --save_folder /path/to/output --start_id 1000

Utility Functions

The script uses the following utility functions from the utils module:

  • utils.bart: Executes BART (Berkeley Advanced Reconstruction Toolbox) commands.
  • utils.check_out: Executes shell commands.
  • utils.getname: Generates a standardized file name for the output.
  • utils.writecfl: Writes complex images to files in CFL format.

Ensure these functions are correctly implemented in the utils module.

Error Handling

The script handles corrupted HDF5 files by printing an error message and skipping the file. If no files are found in the specified data folder, it prints a "No files found" message and exits.

We provided the binary file for BART reconstruction toolbox on this link. If it doesn't work on your local system, please clone the BART repository and compile it on your local system and set the BART_PATH environment variable to the path of the compiled binary.

Contact

For any questions or issues, please contact the script author or maintainer.