Research map / Detection and segmentation / CNN features
Non-Local U-Nets for Biomedical Image Segmentation
Zhengyang Wang, Na Zou, Dinggang Shen, Shuiwang Ji
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 |
|---|---|---|---|
| Replace local-only CC-3D-FCN with non-local U-Net global aggregation blocks | CC-3D-FCN · iSeg infant brain MR, leave-one-subject-out | Average Dice ratio: 0.9087$\pm$0.0066 -> 0.9292$\pm$0.0050 | Improved |
| Replace local-only CC-3D-FCN with non-local U-Net global aggregation blocks | CC-3D-FCN · iSeg infant brain MR, leave-one-subject-out | Average 3D-MHD: 0.4924$\pm$0.0345 -> 0.4319$\pm$0.0313 | Improved |
| Replace CC-3D-FCN with non-local U-Net | CC-3D-FCN · iSeg infant brain MR | Number of parameters: 2,534,276 -> 1,821,124 | Improved |
| Replace CC-3D-FCN with non-local U-Net | CC-3D-FCN · iSeg infant brain MR, patch size $32^3$, overlap step 8 | Inference time (min): 3.85$\pm$0.15 -> 3.06$\pm$0.12 | Improved |
| Add short-range residual connections (ablation) | Model1 · 10th subject, ablation | Average Dice ratio: 0.9103 -> 0.9156 | Improved |
| Add short-range residual connections (ablation) | Model1 · 10th subject, ablation | Average 3D-MHD: 0.4448 -> 0.4312 | Improved |
| Replace the first up-sampling block with an up-sampling global aggregation block (ablation) | Model2 · 10th subject, ablation | Average Dice ratio: 0.9156 -> 0.9174 | Improved |
| Replace the first up-sampling block with an up-sampling global aggregation block (ablation) | Model2 · 10th subject, ablation | Average 3D-MHD: 0.4312 -> 0.4271 | Improved |
| Replace both up-sampling blocks with up-sampling global aggregation blocks (ablation) | Model2 · 10th subject, ablation | Average Dice ratio: 0.9156 -> 0.9186 | Improved |
| Replace both up-sampling blocks with up-sampling global aggregation blocks (ablation) | Model2 · 10th subject, ablation | Average 3D-MHD: 0.4312 -> 0.4220 | Improved |
| Replace the bottom block with a global aggregation block (ablation) | Model2 · 10th subject, ablation | Average Dice ratio: 0.9156 -> 0.9194 | Improved |
| Replace the bottom block with a global aggregation block (ablation) | Model2 · 10th subject, ablation | Average 3D-MHD: 0.4312 -> 0.4211 | Improved |
Closest papers by meaning
- nnWNet: Rethinking the Use of Transformers in Biomedical Image Segmentation and Calling for a Unified Evaluation Benchmark (CVPR 2025 · similarity 0.88)
- U-KAN Makes Strong Backbone for Medical Image Segmentation and Generation (AAAI 2025 · similarity 0.88)
- Upping the Game: How 2D U-Net Skip Connections Flip 3D Segmentation (NeurIPS 2024 · similarity 0.87)
- Rethinking U-Net: Task-Adaptive Mixture of Skip Connections for Enhanced Medical Image Segmentation (AAAI 2025 · similarity 0.87)
- Pancakes: Consistent Multi-Protocol Image Segmentation Across Biomedical Domains (NeurIPS 2025 · similarity 0.85)
- FuseUNet: A Multi-Scale Feature Fusion Method for U-like Networks (ICML 2025 · similarity 0.85)
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.
