Help Privacy Policy Disclaimer
  Advanced SearchBrowse


  Toward sharing brain images: Differentially private TOF-MRA images with segmentation labels using generative adversarial networks

Kossen, T., Hirzel, M. A., Madai, V. I., Boenisch, F., Hennemuth, A., Hildebrand, K., et al. (2022). Toward sharing brain images: Differentially private TOF-MRA images with segmentation labels using generative adversarial networks. Frontiers in Artificial Intelligence, 5: 813842. doi:10.3389/frai.2022.813842.

Item is


show hide
Genre: Journal Article


show Files
hide Files
Kossen_2022.pdf (Publisher version), 2MB
MIME-Type / Checksum:
application/pdf / [MD5]
Technical Metadata:
Copyright Date:
Copyright Info:




Kossen, Tabea1, 2, Author
Hirzel, Manuel A.1, Author
Madai, Vince I.1, 3, 4, Author
Boenisch, Franziska5, Author
Hennemuth, Anja2, 6, 7, Author
Hildebrand, Kristian8, Author
Pokutta, Sebastian9, 10, Author
Sharma, Kartikey9, Author
Hilbert, Adam1, Author
Sobesky, Jan11, 12, Author
Galinovic, Ivana12, Author
Khalil, Ahmed12, 13, 14, Author           
Fiebach, Jochen B.12, Author
Frey, Dietmar1, Author
1Charité Lab for Artificial Intelligence in Medicine (CLAIM), Charité University Medicine Berlin, Germany, ou_persistent22              
2Department of Computer Engineering and Microelectronics, TU Berlin, Germany, ou_persistent22              
3QUEST Center for Responsible Research, Berlin Institute of Health (BIH), Germany, ou_persistent22              
4Faculty of Computing, Engineering and the Built Environment, School of Computing and Digital Technology, Birmingham City University, United Kingdom, ou_persistent22              
5Fraunhofer AISEC, Berlin, Germany, ou_persistent22              
6Institute for Imaging Science and Computational Modelling in Cardiovascular Medicine, Charité University Medicine Berlin, Germany, ou_persistent22              
7Fraunhofer MEVIS, Bremen, Germany, ou_persistent22              
8Department VI Computer Science and Media, Berlin University of Applied Sciences, Germany, ou_persistent22              
9Department for AI in Society, Science, and Technology, Zuse Institute Berlin, Germany, ou_persistent22              
10Institute of Mathematics, TU Berlin, Germany, ou_persistent22              
11Johanna-Etienne-Hospital, Neuss, Germany, ou_persistent22              
12Center for Stroke Research, Charité University Medicine Berlin, Germany, ou_persistent22              
13Department Neurology, MPI for Human Cognitive and Brain Sciences, Max Planck Society, ou_634549              
14MindBrainBody Institute, Berlin School of Mind and Brain, Humboldt University Berlin, Germany, ou_persistent22              


Free keywords: Brain vessel segmentation; Differential privacy; Generative Adversarial Networks; Neuroimaging; Privacy preservation
 Abstract: Sharing labeled data is crucial to acquire large datasets for various Deep Learning applications. In medical imaging, this is often not feasible due to privacy regulations. Whereas anonymization would be a solution, standard techniques have been shown to be partially reversible. Here, synthetic data using a Generative Adversarial Network (GAN) with differential privacy guarantees could be a solution to ensure the patient's privacy while maintaining the predictive properties of the data. In this study, we implemented a Wasserstein GAN (WGAN) with and without differential privacy guarantees to generate privacy-preserving labeled Time-of-Flight Magnetic Resonance Angiography (TOF-MRA) image patches for brain vessel segmentation. The synthesized image-label pairs were used to train a U-net which was evaluated in terms of the segmentation performance on real patient images from two different datasets. Additionally, the Fréchet Inception Distance (FID) was calculated between the generated images and the real images to assess their similarity. During the evaluation using the U-Net and the FID, we explored the effect of different levels of privacy which was represented by the parameter ϵ. With stricter privacy guarantees, the segmentation performance and the similarity to the real patient images in terms of FID decreased. Our best segmentation model, trained on synthetic and private data, achieved a Dice Similarity Coefficient (DSC) of 0.75 for ϵ = 7.4 compared to 0.84 for ϵ = ∞ in a brain vessel segmentation paradigm (DSC of 0.69 and 0.88 on the second test set, respectively). We identified a threshold of ϵ <5 for which the performance (DSC <0.61) became unstable and not usable. Our synthesized labeled TOF-MRA images with strict privacy guarantees retained predictive properties necessary for segmenting the brain vessels. Although further research is warranted regarding generalizability to other imaging modalities and performance improvement, our results mark an encouraging first step for privacy-preserving data sharing in medical imaging.


Language(s): eng - English
 Dates: 2021-11-122022-03-312022-05-02
 Publication Status: Published online
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: DOI: 10.3389/frai.2022.813842
Other: eCollection 2022
PMID: 35586223
PMC: PMC9108458
 Degree: -



Legal Case


Project information

show hide
Project name : -
Grant ID : 777 107
Funding program : -
Funding organization : European Commission
Project name : -
Grant ID : 031B0154
Funding program : -
Funding organization : German Federal Ministry of Education and Research (BMBF)

Source 1

Title: Frontiers in Artificial Intelligence
Source Genre: Journal
Publ. Info: Lausanne, Switzerland : Frontiers Research Foundation
Pages: - Volume / Issue: 5 Sequence Number: 813842 Start / End Page: - Identifier: ISSN: 2624-8212
CoNE: https://pure.mpg.de/cone/journals/resource/2624-8212