HFBSurv: Hierarchical multimodal fusion with factorized bilinear models for cancer survival prediction

This is an implementation of HFBSurv in Python 3.6.13 under Linux with CPU Intel Xeon 4110 @ 2.10GHz, GPU NVIDIA GeForce RTX 2080 Ti, and 192GB of RAM. It follows a modern deep learning design and is implemented by PyTorch platform.

Installation

git clone https://github.com/Liruiqing-ustc/HFBSurv.git

cd HFBSurv

Run the main routine

Examples shown below for training or testing:

Training:

cd model

python train_cv.py

Testing:

cd model

python test_cv.py

Acknowledgments

This code is inspired by SALMON and Pathomic Fusion.