ATHENA: a GPU-based Framework for Biomedical 3D Rigid Image Registration

This repository contains the code relative the publication "ATHENA: a GPU-based Framework for Biomedical 3D Rigid Image Registration" at BioCAS 2023

Testing Environment

1. We tested the code on linux-based machines, on CPUs such as AMD Ryzen 7 5800X, Intel i7-10870H, Intel i7-7700HQ, Intel i7-6700, and Intel Xeon Platinum 8167M; we tested GPUs such as NVIDIA RTX A5000, NVIDIA GTX 1650 Ti, NVIDIA GTX 1050, NVIDIA GTX 960, and NVIDIA Tesla V100.
2. We used python 3.8.10 with pydicom cv2 numpy pandas torch kornia argparse statistics packets
3. Data used in this publication were generated by the National Cancer Institute Clinical Proteomic Tumor Analysis Consortium (CPTAC).https://doi.org/10.7937/k9/tcia.2018.pat12tbs. Patient: C3N-00704, Study: Dec 10, 2000 NM PET 18 FDG SKULL T, CT: WB STND, PET: WB 3D AC)
4. The 1+1 code takes inspiration from ITK code

Code organization

• *.py python source code for the 1+1 or Powell's optimizations procedures, for output evaluation, and for robustness analysis.
• run_script.sh automation script to run extensive tests for both CPU and CUDA-based platforms.
• robustness.sh automation script to run extensive robustness analysis for our CUDA-based platform.

Credits and Contributors

Contributors: Sorrentino, Giuseppe and Venere, Marco and D'Arnese, Eleonora and Conficconi, Davide and Poles, Isabella and Santambrogio, Marco D.

If you find this repository useful, please use the following citation(s):

@inproceedings{sorrentino2023athena,
  title={ATHENA: a GPU-based Framework for Biomedical 3D Rigid Image Registration},
  author={Sorrentino, Giuseppe and Venere, Marco and D'Arnese, Eleonora and Conficconi, Davide and Poles, Isabella and Santambrogio, Marco and others},
  booktitle={2023 IEEE Biomedical Circuits and Systems Conference (BioCAS)},
  pages={1--5},
  year={2023}
}