Research map / Detection and segmentation / Open-set and few-shot detection
DoDNet: Learning To Segment Multi-Organ and Tumors From Multiple Partially Labeled Datasets
Check 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 DoDNet's task-conditioned dynamic segmentation head instead of the strongest listed shared-network baseline | DoDNet · MOTS test | Average Dice: 74.55 -> 75.64 | Helped |
| Use DoDNet's task-conditioned dynamic segmentation head instead of TAL | DoDNet · MOTS test | Average HD: 23.56 -> 19.50 | Helped |
| Pre-train on MOTS rather than train from scratch | BCV fine-tuning · BCV test | Avg. Dice: 85.30 -> 86.44 | Helped |
| Pre-train on MOTS rather than train from scratch | BCV fine-tuning · BCV test | Avg. SD: 1.46 -> 1.17 | Helped |
| Pre-train on MOTS rather than train from scratch | BCV fine-tuning · BCV test | Avg. HD: 19.67 -> 15.62 | Helped |
| Reduce dynamic head depth from 3 to 2 (ablation) | DoDNet · Validation: 11 organs and tumors | Average Dice: 71.67 -> 71.30 | No clear effect |
| Increase dynamic head depth from 3 to 4 (ablation) | DoDNet · Validation: 11 organs and tumors | Average Dice: 71.67 -> 71.63 | No clear effect |
| Reduce dynamic head width from 8 to 4 (ablation) | DoDNet · Validation: 11 organs and tumors | Average Dice: 71.67 -> 69.79 | Hurt |
| Increase dynamic head width from 8 to 16 (ablation) | DoDNet · Validation: 11 organs and tumors | Average Dice: 71.67 -> 71.45 | Hurt |
| Remove image feature conditioning from dynamic filter generation (ablation) | DoDNet · Validation: 11 organs and tumors | Average Dice: 71.67 -> 71.26 | Hurt |
| Remove image feature conditioning from dynamic filter generation (ablation) | DoDNet · Validation: 11 organs and tumors | Average HD: 25.86 -> 29.38 | Hurt |
| Remove task encoding from dynamic filter generation (ablation) | DoDNet · Validation: 11 organs and tumors | Average Dice: 71.67 -> 51.80 | Hurt |
Closest papers by meaning
- CCQ: Cross-Class Query Network for Partially Labeled Organ Segmentation (AAAI 2023 · similarity 0.90)
- CAT: Coordinating Anatomical-Textual Prompts for Multi-Organ and Tumor Segmentation (NeurIPS 2024 · similarity 0.89)
- nnWNet: Rethinking the Use of Transformers in Biomedical Image Segmentation and Calling for a Unified Evaluation Benchmark (CVPR 2025 · similarity 0.87)
- SGTC: Semantic-Guided Triplet Co-training for Sparsely Annotated Semi-Supervised Medical Image Segmentation (AAAI 2025 · similarity 0.86)
- SuperLightNet: Lightweight Parameter Aggregation Network for Multimodal Brain Tumor Segmentation (CVPR 2025 · similarity 0.86)
- E2ENet: Dynamic Sparse Feature Fusion for Accurate and Efficient 3D Medical Image Segmentation (NeurIPS 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.