Research map / Vision-language and multimodal / Cross-modal learning
M3AE: Multimodal Representation Learning for Brain Tumor Segmentation with Missing Modalities
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use M3AE for missing-modality segmentation | BraTS 2018 test · BraTS 2018, all modality combinations | Average whole-tumor DSC (%): 85.5 -> 85.8 | Helped |
| Use M3AE for missing-modality segmentation | BraTS 2018 test · BraTS 2018, all modality combinations | Average tumor-core DSC (%): 76.0 -> 77.4 | Helped |
| Use M3AE for missing-modality segmentation | BraTS 2018 test · BraTS 2018, all modality combinations | Average enhancing-tumor DSC (%): 56.6 -> 59.9 | Helped |
| Use M3AE for missing-modality segmentation | BraTS 2020 test · BraTS 2020, all modality combinations | Whole-tumor DSC (%): 86.7 -> 86.9 | Helped |
| Use M3AE for missing-modality segmentation | BraTS 2020 test · BraTS 2020, all modality combinations | Tumor-core DSC (%): 78.2 -> 79.1 | Helped |
| Use M3AE for missing-modality segmentation | BraTS 2020 test · BraTS 2020, all modality combinations | Enhancing-tumor DSC (%): 59.7 -> 61.7 | Helped |
| Add M3AE self-supervised pretraining (ablation) | BraTS 2018 validation segmentation · BraTS 2018 validation | Average DSC (%): 72.9 -> 75.9 | Helped |
| Add patch masking to modality-dropout pretraining (ablation) | M3AE · BraTS 2018 validation | Average DSC (%): 74.8 -> 75.9 | Helped |
| Use optimized model-inversion substitute instead of zero filling (ablation) | M3AE · BraTS 2018 validation | Average DSC (%): 74.4 -> 75.9 | Helped |
| Use optimized model-inversion substitute instead of the training-set mean image (ablation) | M3AE · BraTS 2018 validation | Average DSC (%): 45.2 -> 75.9 | Helped |
| Add heterogeneous missing-modal self-distillation (ablation) | M3AE · BraTS 2018 validation | Average DSC (%): 75.1 -> 75.9 | Helped |
Closest papers by meaning
- Semantic-guided Masked Mutual Learning for Multi-modal Brain Tumor Segmentation with Arbitrary Missing Modalities (AAAI 2025 · similarity 0.89)
- Incomplete Multi-modal Brain Tumor Segmentation via Learnable Sorting State Space Model (CVPR 2025 · similarity 0.87)
- KMD: Koopman Multi-modality Decomposition for Generalized Brain Tumor Segmentation under Incomplete Modalities (CVPR 2025 · similarity 0.87)
- UniMRSeg: Unified Modality-Relax Segmentation via Hierarchical Self-Supervised Compensation (NeurIPS 2025 · similarity 0.86)
- TMFormer: Token Merging Transformer for Brain Tumor Segmentation with Missing Modalities (AAAI 2024 · similarity 0.86)
- NaMa: Neighbor-Aware Multi-Modal Adaptive Learning for Prostate Tumor Segmentation on Anisotropic MR Images (AAAI 2024 · 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.