Virtual Nodes Guided Dynamic Graph Neural Network for Brain Tumor Segmentation with Missing Modalities
A new study presents a graph-based framework for brain tumor segmentation that addresses the common issue of missing modalities in MRI data. The proposed method utilizes virtual nodes to enhance the model's robustness and dynamically adjusts connections based on available modalities. Experimental results indicate that this approach outperforms existing state-of-the-art methods in handling incomplete modality scenarios.
- ▪The study introduces a one-stage framework for brain tumor segmentation using a dynamic graph neural network.
- ▪Virtual nodes are used to provide supplementary information for missing modalities.
- ▪The method shows improved performance on the BRATS-2018 and BRATS-2020 datasets compared to existing techniques.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.16880 (cs) [Submitted on 16 May 2026] Title:Virtual Nodes Guided Dynamic Graph Neural Network for Brain Tumor Segmentation with Missing Modalities Authors:Sha Tao, Jiao Pan, Yu Guo, Chao Yao View a PDF of the paper titled Virtual Nodes Guided Dynamic Graph Neural Network for Brain Tumor Segmentation with Missing Modalities, by Sha Tao and 3 other authors View PDF HTML (experimental) Abstract:Multimodal magnetic resonance imaging (MRI) is crucial for brain tumor segmentation, with many methods leveraging its four key modalities to capture complementary information for effective sub-region analysis.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
