Tumor Synthesis conditioned on Radiomics

Tumor Synthesis conditioned on Radiomics [Jonghun Kim], [Inye Na], [Eunsook ko], [Hyunjin Park]

Overview

This repository contains the code for Tumor Synthesis conditioned on Radiomics.

The model architecture is illustrated below:

Our code was written by applying VQ-GAN, GigaGAN and BBDM into 3D methods.

We would like to thank those who have shared their code.

Thanks to everyone who contributed this works and codes.

• Taming Transformers for High-Resolution Image Synthesis

• Scaling up GANs for Text-to-Image Synthesis

• BBDM: Image-to-image Translation with Brownian Bridge Diffusion Models

Feature Extraction

We used pyradiomics to extract radiomics features.

For more details on pyradiomics, please refer to the following document.

https://pyradiomics.readthedocs.io/en/latest/index.html

Citation

@InProceedings{Kim_2025_WACV,
    author    = {Kim, Jonghun and Na, Inye and Ko, Eun Sook and Park, Hyunjin},
    title     = {Tumor Synthesis Conditioned on Radiomics},
    booktitle = {Proceedings of the Winter Conference on Applications of Computer Vision (WACV)},
    month     = {February},
    year      = {2025},
    pages     = {3635-3646}
}