Towards Graph Representation Learning Based Surgical Workflow Anticipation

Example Code for the paper Towards Graph Representation Learning Based Surgical Workflow Anticipation

Environment Setup

First please create an appropriate environment using conda:

conda env create -f surgery.yaml

conda activate surgery

Test Pre-Trained Models

Evaluate on Sample dataset:

python main_infer.py

Train a Model

Train on Sample dataset:

python main.py

In training, our default stream is only based on graph-level information, as we proposed in our paper.

Citing

If you find this work useful, please consider our paper to cite:

@inproceedings{zhang22towards,
 author={Zhang, Francis Xiatian and Moubayed, Noura Al and Shum, Hubert P. H.},
 booktitle={Proceedings of the 2022 IEEE-EMBS International Conference on Biomedical and Health Informatics},
 title={Towards Graph Representation Learning Based Surgical Workflow Anticipation },
 year={2022},
 publisher={IEEE},
 location={Ioannina, Greece},
}