A Novel Multi-modal Population-graph based Framework for Patients of Esophageal Squamous Cell Cancer Prognostic Risk Prediction
- Linux (tested on Ubuntu 16.04, 18.04, 20.04)
- Python 3.7+
- PyTorch 1.6 or higher (tested on PyTorch 1.13.1)
- CUDA 11.3 or higher (tested on CUDA 11.6)
conda env create -f environment.yml
The training and evaluation code of stage 1 and 2 can be overviewed in main_stage1.py and main_stage2.py. The code of proposed MPGSurv model can be seen in /models.
we apologize that we do not have the right to disclose the datasets collected by the hospital. Due to ethical review and patient privacy disclosure restrictions, we are unable to request public disclosure of the dataset at this time. In addition, future work will be based on this dataset and it is not appropriate to disclose the dataset at this time. The data types and requirements can be set according to /dataloader.
Our repo is developed based on the these projects: MMMNA-Net, GraphSAGE