A Novel Digital Score for Abundance of Tumour Infiltrating Lymphocytes Predicts Disease Free Survival in Oral Squamous Cell Carcinoma
This repository contains the implementation of TILAb-score as described in the paper.
The journal paper on this work has been published in Nature Scientific Reports. If you use this code in your research, please cite this work:
@article{shaban2019novel,
title={A novel Digital Score for Abundance of Tumour Infiltrating Lymphocytes predicts Disease free Survival in oral Squamous cell carcinoma},
author={Shaban, Muhammad and Khurram, Syed Ali and Fraz, Muhammad Moazam and Alsubaie, Najah and Masood, Iqra and Mushtaq, Sajid and Hassan, Mariam and Loya, Asif and Rajpoot, Nasir M},
journal={Scientific reports},
volume={9},
number={1},
pages={1--13},
year={2019},
publisher={Nature Publishing Group}
}
The datset for training should be organized in following hierarchy:
dataset
-- train
-- 0_Stroma
-- 1_Non_ROI
-- 2_Tumour
-- 3_Lymphocyte
-- valid
-- 0_Stroma
-- 1_Non_ROI
-- 2_Tumour
-- 3_Lymphocyte
Please contact Prof. Nasir Rajpoot ([email protected]) for dataset related queries.
The training.py file in src/
directory will train the model using the dataset in dataset/
directory. You may need to tune the hyperparameters for training on your own dataset to train an optimal model.
The trained model used to produce the results in the paper is available in the models/
directory.
Following software packages will be required to run this code:
-- Python 3.5
-- tensorflow-gpu=1.8.0
-- keras=2.1.6
-- openslide
-- opencv_python
-- scipy
-- R packages
-- survival
-- survMisc
-- gdata
-- ggplot2
-- survminer
-- rms
See the list of contributors who participated in this project.
This project is licensed under the GNU General Public License - see the LICENSE.md file for details.