Histogram-Standardization

This repository contains the NiftyNet implementation of histogram standardization¹, isolated from NiftyNet for convenience.

Installation instructions

1. Clone/download repository.

2. Change directory into repository.

3. Install:

   pip install .
   

How to use it

• Function imports.

• Command line:

  • Training:

     hist_train --train_dir .\sample_data\0_sample_train_t2w --landmarks_dir .\sample_data\0_sample_train_t2w
    

  • Inference:

     hist_std --infer_dir .\sample_data\0_sample_test_t2w --output_dir .\sample_data\1_standardized_test_t2w --landmarks_path .\sample_data\0_sample_train_t2w\landmarks.txt
    

References

¹ Nyúl, L.G.; Udupa, J.K.; Zhang, X. New variants of a method of MRI scale standardization. IEEE Trans. Med. Imaging 2000, 19, 143–150, doi:10.1109/42.836373