Thanks to visit codestin.com
Credit goes to github.com

Skip to content

POLARIS-Sheffield/1H-MRI-segmentation

Repository files navigation

1HMRI-segmentation

Using the DL trained model

The model was trained using the Niftynet framework; instructions are provided below to allow researchers to utilize the trained DL model. The instructions given here will largely follow the instructions given at https://github.com/NifTK/NiftyNet.

Installation

pip install “tensorflow==1.15.*”
pip install niftynet

We have provided a configuration file that can be used for inference on 1H MRI scans. We highlight the main parameters in the configuration file required for inference that are specific to the primary user.

Paths

The user must provide paths to folders containing the 1H MRI scans and a path to the location to save the outputted segmentations.


[Proton]

path_to_search = #path to 1H MRI scans
save_seg_dir = #path to output location

Computational resources

Depending on the computational resources available to the individual researcher, the following parameters will need to be amended:

num_threads = 0
num_gpus = 0

In our investigations, we used a single GPU with two threads.

Running inference

The following command can be used to run inference by indicating the path to the configuration file:

net_segment inference -c /path_to_config/config_inference.ini

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published