Skip to content

This repositories include python code for infant deep learning segmentation framework we developed.

Notifications You must be signed in to change notification settings

wangyunai/ID-Seg

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ID-Seg

This repositories include python code for infant deep learning segmentation framework we developed.

Training steps

First, we trained our model on a big dataset (dhcp dataset) using pretrain script on three main plain views (axial, coronal, and sagittal) and tested the performance of our 3D model using pretrain_validation code.

After that, we used train script for transfering our knowledge on a smaller dataset (MCRIB), where we used Leave One Out Cross Validation (LOOCV) technique. Again, we trained our model on three different plain views. Using save_prediction code, we saved predictions of our model on our dataset and then used test script to calculate accuracy and dice scores of results.

How to use

The codes are ready to use. You just need to install needed requirements and provide the paths to dataset as instructed by comments in the code. Requirements are nibabel, scikit-image, scikit-learn and pytorch, nn-common-modules.

Refrences

quickNAT_pytorch

About

This repositories include python code for infant deep learning segmentation framework we developed.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages