README

This is the official Pytorch implementation of LoTeNet model in "Tensor Networks for Medical Image Classification", Raghavendra Selvan & Erik Dam, MIDL 2020. Runner up for the best paper award at MIDL2020.

What is this repository for?

• Run and reproduce results in the paper on LIDC dataset
• v1.0

How do I get set up?

• Basic Pytorch dependency
• Tested on Pytorch 1.3, Python 3.6
• Download the data from here
• Unzip the data and point the path to --data_path
• How to run tests: python train.py --data_path data_location

Usage guidelines

• Kindly cite our publication if you use any part of the code

@inproceedings{
raghav2020tensor,
title={Tensor Networks for Medical Image Classification},
author={Raghavendra Selvan, Erik B Dam},
booktitle={International Conference on Medical Imaging with Deep Learning -- Full Paper Track},
year={2020},
month={July},
url={https://openreview.net/forum?id=jjk6bxk07G}}

Who do I talk to?

• raghav@di.ku.dk

Thanks to the following repositories we base our project on:

• Torch MPS for the amazing MPS in Pytorch implementations
• Prob.U-Net for preprocessing LIDC data
• Dense Net implementation