Skip to content

liviamarodrigues/hsynex

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

H-SynEx

A tool for the hypothalamus and its subregions capable of working across different MRI modalities and resolution. H-SynEx was trained on synthetic images derived from high-resolution ex vivo MRI.

The model was trained on synthetic images derived from ex vivo MRI label maps:

Example of testing images:

Labels

Label Subregion
1 left anterior inferior
2 left posterior
3 left tuberal inferior
4 left tuberal superior
6 left anterior superior
7 right anterior inferior
8 right posterior
9 right tuberal inferior
10 right tuberal superior
12 right anterior superior

USAGE

To use the tool, please follow the following steps:

  • Download this repository

  • Using the terminal, enter the inference folder:

     cd <your_path>/hsynex/inference
    
  • Run the 'find_hypothalamus_subnuclei.py':

      python --input_path <input_path> --out_path: <out_path>
    

Where:

  • <input_path> is the path where the input MR images are located
  • <out_path> is the chosen path to save the segmentations

DEPENDENCIES

This code was implemented using Python 3.10. To run it, install the dependencies listed at requirements.txt.

This repository also relies on a modified implementation of pytorch-3dunet, which is already included.

Finally, it is necessary to install FreeSurfer 7.4

CITATION

If you use H-SynEx in your project, please cite:

[arXiv]: Rodrigues, Livia, et al. "H-SynEx: Using synthetic images and ultra-high resolution ex vivo MRI for hypothalamus subregion segmentation." arXiv preprint arXiv:2401.17104 (2024).

About

A tool for hypothalamus subregion segmentation capable of working across different MRI modalities and resolution

Resources

License

Stars

Watchers

Forks

Packages

No packages published