PIANO: Perfusion Imaging via Advection-Diffusion

Please refer to the newest version at https://github.com/uncbiag/D2-SONATA.

This repository contains the code and experiments for our papers:

Perfusion Imaging: An Advection Diffusion Approach - IEEE TMI (In Minor Revision)

PIANO: Perfusion Imaging via Advection-Diffusion - MICCAI 2020 (Early Accept)

Environment

Training and evaluation environment: Python 3.11.4, PyTorch 2.0.1, CUDA 12.2. Run the following command to install the required packages.

conda create -n piano python=3.11
conda activate piano

git clone https://github.com/uncbiag/PIANO
cd /path/to/piano
pip install -r requirements.txt

2D Demo

cd learning
python demo_2d.py

Train on Your Own Perfusion Dataset

cd learning
python learn_all_MRP.py 

Cite this work

@misc{liu2020perfusion,
  title={Perfusion Imaging: An Advection Diffusion Approach}, 
  author={Peirong Liu and Yueh Z. Lee and Stephen R. Aylward and Marc Niethammer},
  year={2020},
  eprint={2009.02796},
  archivePrefix={arXiv},
  primaryClass={eess.IV}
}

@InProceedings{liu2020piano,
  title="{PIANO}: Perfusion Imaging via Advection-Diffusion",
  author={Peirong Liu and Yueh Z. Lee and Stephen R. Aylward and Marc Niethammer}, 
  booktitle={Medical Image Computing and Computer Assisted Intervention (MICCAI)},
}