This is the implementation of our submitted Computers in Biology and Medicine 2024 work 'MITS-GAN: Safeguarding Medical Imaging from Tampering with Generative Adversarial Networks'. This study introduces MITS-GAN, a novel approach to prevent tampering in medical images, with a specific focus on CT scans. The approach disrupts the output of the attacker's CT-GAN architecture by introducing imperceptible but yet precise perturbations. The original paper can be found here.
Please leave a star ⭐ and cite the following paper if you use this repository for your project.
Models and code will be publicly available after the paper publication.

@misc{pasqualino2024mitsgan,
      title={MITS-GAN: Safeguarding Medical Imaging from Tampering with Generative Adversarial Networks}, 
      author={Giovanni Pasqualino and Luca Guarnera and Alessandro Ortis and Sebastiano Battiato},
      year={2024},
      eprint={2401.09624},
      archivePrefix={arXiv},
      primaryClass={eess.IV}
}

Choose one of the two installation options:

Quickstart here 👉 [

Install the required dependencies by running the following command in your terminal:

pip install --upgrade scipy matplotlib pandas tensorflow keras SimpleITK pydicom torch

Dataset is available here

Qualitative results on the reconstruction task compared with images as manipulation targets.

Metric results evaluated between the following pairs on the: real-MITS-GAN, real-TAFIM, real-MITS-GAN tampered and real-TAFIM tampered. Lower values are better for RMSE and LPIPS, higher for PSNR.