Search: L773:2052 4463
> (2024) >
Brain tumor segment...
Brain tumor segmentation using synthetic MR images - A comparison of GANs and diffusion models
-
- Akbar, Muhammad Usman, 1990- (author)
- Linköpings universitet,Avdelningen för medicinsk teknik,Tekniska fakulteten,Centrum för medicinsk bildvetenskap och visualisering, CMIV
-
- Larsson, Måns (author)
- Eigenvision, Malmö, Sweden
-
- Blystad, Ida, 1972- (author)
- Linköpings universitet,Medicinska fakulteten,Centrum för medicinsk bildvetenskap och visualisering, CMIV,Avdelningen för diagnostik och specialistmedicin,Region Östergötland, Röntgenkliniken i Linköping
-
show more...
-
- Eklund, Anders, 1981- (author)
- Linköpings universitet,Avdelningen för medicinsk teknik,Tekniska fakulteten,Centrum för medicinsk bildvetenskap och visualisering, CMIV,Statistik och maskininlärning
-
show less...
-
(creator_code:org_t)
- Nature Publishing Group, 2024
- 2024
- English.
-
In: Scientific Data. - : Nature Publishing Group. - 2052-4463. ; 11:1
- Related links:
-
https://doi.org/10.1...
-
show more...
-
https://liu.diva-por... (primary) (Raw object)
-
https://urn.kb.se/re...
-
https://doi.org/10.1...
-
show less...
Abstract
Subject headings
Close
- Large annotated datasets are required for training deep learning models, but in medical imaging data sharing is often complicated due to ethics, anonymization and data protection legislation. Generative AI models, such as generative adversarial networks (GANs) and diffusion models, can today produce very realistic synthetic images, and can potentially facilitate data sharing. However, in order to share synthetic medical images it must first be demonstrated that they can be used for training different networks with acceptable performance. Here, we therefore comprehensively evaluate four GANs (progressive GAN, StyleGAN 1–3) and a diffusion model for the task of brain tumor segmentation (using two segmentation networks, U-Net and a Swin transformer). Our results show that segmentation networks trained on synthetic images reach Dice scores that are 80%–90% of Dice scores when training with real images, but that memorization of the training images can be a problem for diffusion models if the original dataset is too small. Our conclusion is that sharing synthetic medical images is a viable option to sharing real images, but that further work is required. The trained generative models and the generated synthetic images are shared on AIDA data hub.
Subject headings
- MEDICIN OCH HÄLSOVETENSKAP -- Klinisk medicin -- Radiologi och bildbehandling (hsv//swe)
- MEDICAL AND HEALTH SCIENCES -- Clinical Medicine -- Radiology, Nuclear Medicine and Medical Imaging (hsv//eng)
- TEKNIK OCH TEKNOLOGIER -- Medicinteknik -- Medicinsk bildbehandling (hsv//swe)
- ENGINEERING AND TECHNOLOGY -- Medical Engineering -- Medical Image Processing (hsv//eng)
Keyword
- Deep learning
- brain tumor
- magnetic resonance imaging
- synthetic images
- generative adversarial networks
- diffusion models
Publication and Content Type
- ref (subject category)
- art (subject category)
Find in a library
To the university's database