SDCT-AuxNet^${\theta}$:DCT Augmented Stain Deconvolutional CNN with-Auxiliary Classifier for Cancer Diagnosis

The implementation of the paper SDCT-AuxNet$^{\theta}$: DCT Augmented Stain Deconvolutional CNN with Auxiliary Classifier for Cancer Diagnosis. The article is available here. The preprint of the article is also available here.

The code has been tested with Python 2.7 and PyTorch 0.4.1

Dataset:

The dataset for this article is available at The Cancer Imaging Archive (TCIA)

Training

1. For training the the dataset directory should have the following structure:

main_data_dir
|          -----------class1------subject_folders
|         |
fold0------
|         |
|          -----------class2------subject_folders
| 
|          -----------class1------subject_folders
|         |
fold1------
|         |
|          -----------class2------subject_folders 
|
|
.
.          -----------class1------subject_folders
.         |
foldn------
          |
           -----------class2------subject_folders

2. Run train_model.py

Evaluation

1. Run test_model.py
2. The performance of the model on the test set can be evaluated at the challenge portal.