Research map / Vision-language and multimodal / Cross-modal learning
Tackling Dual-stage Missing Modalities in Brain Tumor Segmentation via Robust Modality Reconstruction and Prompt-guided Modality Adaptation
Yunpeng Zhao, Cheng Chen, Qing You Pang, Yibing Fu, Quanzheng Li, Carol Tang, Beng Ti Ang, Yueming Jin
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add distribution approximation (segmentation-loss supervision of reconstructed modalities) in pre-training (ablation) | Mask-Rec baseline · BraTS2018 (1% FM) | DSC avg: 72.01 -> 73.44 | Improved |
| add prompt-guided modality adaptation (ablation) | Mask-Rec + DA · BraTS2018 (1% FM) | DSC avg: 73.44 -> 74.32 | Improved |
| add complete-then-distill (ablation) | Mask-Rec + DA + PGMA · BraTS2018 (1% FM) | DSC avg: 74.32 -> 74.57 | Improved |
| add distribution approximation under full-modal training (ablation) | Mask-Rec baseline · BraTS2018 (100% FM) | DSC avg: 74.20 -> 74.72 | Improved |
| distill from teacher using only available real modalities (ablation) | Mask-Rec + DA · BraTS2018 (1% FM) | DSC avg: 73.44 -> 73.41 | No clear effect |
| complete missing modalities with Mrec before distillation (ablation) | Mask-Rec + DA · BraTS2018 (1% FM) | DSC avg: 73.41 -> 73.87 | Improved |
| continually refine Mrec with segmentation loss during CTD (ablation) | CTD · BraTS2018 (1% FM) | DSC avg: 73.87 -> 74.25 | Improved |
| use CLIP text prompt plus visual features instead of binary modality code for hyper-network (ablation) | PGMA · BraTS2018 (1% FM) | DSC avg: 73.82 -> 74.32 | Improved |
| use binary-indicator hyper-network adaptation (ablation) | Mask-Rec + DA · BraTS2018 (1% FM) | DSC avg: 73.44 -> 73.82 | Improved |
| method vs M3AE under 1% FM | proposed method · BraTS2018 (1% FM) | DSC ET: 55.32 -> 60.16 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- M3AE — M3AE: Multimodal Representation Learning for Brain Tumor Segmentation with Missing Modalities (AAAI 2023) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Semantic-guided Masked Mutual Learning for Multi-modal Brain Tumor Segmentation with Arbitrary Missing Modalities (AAAI 2025 · similarity 0.91)
- UniMRSeg: Unified Modality-Relax Segmentation via Hierarchical Self-Supervised Compensation (NeurIPS 2025 · similarity 0.90)
- Incomplete Multi-modal Brain Tumor Segmentation via Learnable Sorting State Space Model (CVPR 2025 · similarity 0.88)
- KMD: Koopman Multi-modality Decomposition for Generalized Brain Tumor Segmentation under Incomplete Modalities (CVPR 2025 · similarity 0.88)
- TMFormer: Token Merging Transformer for Brain Tumor Segmentation with Missing Modalities (AAAI 2024 · similarity 0.86)
- Distilled Prompt Learning for Incomplete Multimodal Survival Prediction (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
