Skip to content

hjahan58/DUAL-energy-CT

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ImageDecomposition-DECT

This reposity is organized mainly for an image decomposition algorithm which is proposed to solve the material decomposition problem in Dual-energy Computed Tomography (DECT).

The algorithm is designed based on deep learning paradigm. For more theoretical details, please go to Deep Learning and Material Decomposition Using DECT.

The algorithm is related to the paper "Image Decomposition Algorithm for Dual-Energy Computed Tomography via Fully Convolutional Network". (DOI: 10.1155/2018/2527516)

All have been tested with python 3.6 and tensorflow 1.4.0 in Linux.

  • checkpoint: the checkpoint path for the model trained with tensorflow. The pre-trained model was trained on a dataset which contained totally 2,454,300 samples. Each sample is a 65*65 image patch extracted from 5987 image slices.
  • data: contains 2 path.
    • test: two test data files, 'test_data_cranial.mat' and 'test_data_pleural.mat'.
    • train: we only provide a sub-set (90,000 training samples) in the 'training_samples_90000.rar' file which can be download from here.
  • result: save decomposition result.
  • src: the codes for three decomposition algorithms:
    • Direct matrix inversion (matrix_inversion.m)
    • Iterative decomposition (iterative_decomposition.m). Related paper: Iterative image-domain decomposition for dual-energy CT
    • The proposed deep model (main.py). After download the pre-trained mode, you can use the following command to run the algorithm.

    " python main.py --dataset="../data/test/test_data_crainal.mat" --model="feedforward" --model_name="your-saved-result-name" --checkpoint="../checkpoint/FCN_trained_model "

Contact

Email: [email protected]

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published