Anonymization of labeled TOF MRA images for brain vessel segmentation using generative adversarial networks

Anonymization and data sharing are crucial for privacy protection and acquisition of large datasets for medical image analysis . Here, the brain’s unique structure allows for re-identification and thus requires non-conventional anonymization . Generative adversarial networks (GANs) have the potential to provide anonymous images while preserving predictive properties . The WGAN-GP-SN showed the highest performance to predict vessels (DSC/95HD 0.82/28.97) benchmarked by the U-net trained on real data . The transfer learning approach showed superior performance for the same GAN compared to no pre-training, especially for one patient only (0.91/25.68 vs. 0.85/27.36) In this work, synthetic image-label pairs retained generalizable information and showed good performance for vessel segmentation. Besides, we showed that synthetic patches can be used in a transfer learning method with independent data. This paves the way to overcome the challenges of scarce data and anonymization in medical imaging, imaging, researchers say. The U-Net trained on synthetic data generated by the GAN showed the best performance to predictions by the WGAN. SN showed the most successful GAN, compared to 0.89/26.61 (Discisciscarencencencerer of 0.93/

Links: PDF - Abstract

Code :

None

Keywords : showed - data - performance - anonymization - synthetic -

Leave a Reply

Your email address will not be published. Required fields are marked *