Beware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray images
Diffusion models were initially developed for text-to-image generation and are now being utilized to generate high quality synthetic images. Preceded by generative adversarial networks (GANs), diffusion models have shown impressive results using various evaluation metrics. However, commonly used met...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2025-01-01
|
| Series: | Machine Learning: Science and Technology |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/2632-2153/ad9a3a |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832582659497263104 |
|---|---|
| author | Muhammad Usman Akbar Wuhao Wang Anders Eklund |
| author_facet | Muhammad Usman Akbar Wuhao Wang Anders Eklund |
| author_sort | Muhammad Usman Akbar |
| collection | DOAJ |
| description | Diffusion models were initially developed for text-to-image generation and are now being utilized to generate high quality synthetic images. Preceded by generative adversarial networks (GANs), diffusion models have shown impressive results using various evaluation metrics. However, commonly used metrics such as Frechet inception distance and inception score are not suitable for determining whether diffusion models are simply reproducing the training images. Here we train StyleGAN and a diffusion model, using BRATS20, BRATS21 and a chest x-ray (CXR) pneumonia dataset, to synthesize brain MRI and CXR images, and measure the correlation between the synthetic images and all training images. Our results show that diffusion models are more likely to memorize the training images, compared to StyleGAN, especially for small datasets and when using 2D slices from 3D volumes. Researchers should be careful when using diffusion models (and to some extent GANs) for medical imaging, if the final goal is to share the synthetic images. |
| format | Article |
| id | doaj-art-02398ecb650042dbbc9cfc612ddecb02 |
| institution | Kabale University |
| issn | 2632-2153 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Machine Learning: Science and Technology |
| spelling | doaj-art-02398ecb650042dbbc9cfc612ddecb022025-01-29T11:27:29ZengIOP PublishingMachine Learning: Science and Technology2632-21532025-01-016101502210.1088/2632-2153/ad9a3aBeware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray imagesMuhammad Usman Akbar0https://orcid.org/0000-0002-3248-5132Wuhao Wang1Anders Eklund2https://orcid.org/0000-0001-7061-7995Department of Biomedical Engineering, Linköping University , Linköping, Sweden; Center for Medical Image Science and Visualization (CMIV), Linköping University , Linköping, Sweden; Clinical Memory Research Unit, Department of Clinical Sciences, Lund University , Malmö, SwedenDivision of Statistics & Machine Learning, Department of Computer and Information Science, Linköping University , Linköping, SwedenDepartment of Biomedical Engineering, Linköping University , Linköping, Sweden; Division of Statistics & Machine Learning, Department of Computer and Information Science, Linköping University , Linköping, Sweden; Center for Medical Image Science and Visualization (CMIV), Linköping University , Linköping, SwedenDiffusion models were initially developed for text-to-image generation and are now being utilized to generate high quality synthetic images. Preceded by generative adversarial networks (GANs), diffusion models have shown impressive results using various evaluation metrics. However, commonly used metrics such as Frechet inception distance and inception score are not suitable for determining whether diffusion models are simply reproducing the training images. Here we train StyleGAN and a diffusion model, using BRATS20, BRATS21 and a chest x-ray (CXR) pneumonia dataset, to synthesize brain MRI and CXR images, and measure the correlation between the synthetic images and all training images. Our results show that diffusion models are more likely to memorize the training images, compared to StyleGAN, especially for small datasets and when using 2D slices from 3D volumes. Researchers should be careful when using diffusion models (and to some extent GANs) for medical imaging, if the final goal is to share the synthetic images.https://doi.org/10.1088/2632-2153/ad9a3abrain MRICXRgenerative AIGANsdiffusion modelsynthetic data |
| spellingShingle | Muhammad Usman Akbar Wuhao Wang Anders Eklund Beware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray images Machine Learning: Science and Technology brain MRI CXR generative AI GANs diffusion model synthetic data |
| title | Beware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray images |
| title_full | Beware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray images |
| title_fullStr | Beware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray images |
| title_full_unstemmed | Beware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray images |
| title_short | Beware of diffusion models for synthesizing medical images—a comparison with GANs in terms of memorizing brain MRI and chest x-ray images |
| title_sort | beware of diffusion models for synthesizing medical images a comparison with gans in terms of memorizing brain mri and chest x ray images |
| topic | brain MRI CXR generative AI GANs diffusion model synthetic data |
| url | https://doi.org/10.1088/2632-2153/ad9a3a |
| work_keys_str_mv | AT muhammadusmanakbar bewareofdiffusionmodelsforsynthesizingmedicalimagesacomparisonwithgansintermsofmemorizingbrainmriandchestxrayimages AT wuhaowang bewareofdiffusionmodelsforsynthesizingmedicalimagesacomparisonwithgansintermsofmemorizingbrainmriandchestxrayimages AT anderseklund bewareofdiffusionmodelsforsynthesizingmedicalimagesacomparisonwithgansintermsofmemorizingbrainmriandchestxrayimages |