EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

A Multi‐Sequence Adversarial Fusion U‐Net for Brain Tumor Image Segmentation

Jie Wang, Jinglu Hu

IEEJ Transactions on Electrical and Electronic Engineering · 05.07.2026

DOI: 10.1002/tee.70376 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEJ Transactions on Electrical and Electronic Engineering, online erschienen am 05.07.2026. Verfasst von Jie Wang und 1 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In the field of brain tumor image segmentation, in order to avoid the impact of insufficient number of training samples, the method of fusing multi‐modal MRI information before segmentation is widely used. However, when fusing different modal features, existing methods only add fixed weights to the features of each modality, resulting in insufficient adaptability to different tasks. To address this issue, this paper proposes a multi‐sequence adversarial fusion U‐Net for automatic segmentation of brain tumor images. Specifically, we design an adversarial neural network (FFN) for multi‐modal feature fusion to map features of different modalities to feature space to generate fused features. At the same time, a multi‐class discriminator (MCD) is used to identify the generated fused features. FFN and MCD perform adversarial learning until MCD cannot distinguish which modality the fused features come from, thereby achieving comprehensive feature fusion across multiple MRI sequences. In this way, we implement an effective modality selection mechanism that allows the model to adaptively generate dynamic weights for features of different modalities. The proposed model is evaluated on the BRATS 2018 dataset, and our model outperforms other multi‐modal fusion segmentation models. In addition, ablation experiments verify the effectiveness of each module in MSAF‐UNet. © 2026 The Author(s). IEEJ Transactions on Electrical and Electronic Engineering published by Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.

Zitieren

APA

Wang, J., & Hu, J. (2026). A Multi‐Sequence Adversarial Fusion U‐Net for Brain Tumor Image Segmentation. IEEJ Transactions on Electrical and Electronic Engineering. https://doi.org/10.1002/tee.70376

Vancouver

Wang J, Hu J. A Multi‐Sequence Adversarial Fusion U‐Net for Brain Tumor Image Segmentation. IEEJ Transactions on Electrical and Electronic Engineering. 2026. doi:10.1002/tee.70376

BibTeX
@article{Wang2026,
  author = {Wang, Jie and Hu, Jinglu},
  title = {A Multi‐Sequence Adversarial Fusion U‐Net for Brain Tumor Image Segmentation},
  journal = {IEEJ Transactions on Electrical and Electronic Engineering},
  year = {2026},
  doi = {10.1002/tee.70376}
}
Kostenlos starten →