Skip to content

Latest commit

 

History

History
34 lines (22 loc) · 881 Bytes

readme.md

File metadata and controls

34 lines (22 loc) · 881 Bytes

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