Research map / Detection and segmentation / CNN features
Difficulty-Aware Attention Network with Confidence Learning for Medical Image Segmentation
Dong Nie, Li Wang, Lei Xiang, Sihang Zhou, Ehsan Adeli, Dinggang Shen
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 difficulty-aware attention mechanism (ablation) | enUNet · Pelvic prostate base | DSC (%): 86.70 -> 89.41 | Improved |
| add difficulty-aware attention mechanism (ablation) | enUNet · Pelvic prostate middle | DSC (%): 87.91 -> 90.90 | Improved |
| add difficulty-aware attention mechanism (ablation) | enUNet · Pelvic prostate apex | DSC (%): 83.92 -> 88.21 | Improved |
| use focal loss (ablation) | enUNet · Pelvic prostate apex | DSC (%): 83.92 -> 86.83 | Improved |
| use focal loss (ablation) | enUNet · Pelvic prostate base | DSC (%): 86.70 -> 88.24 | Improved |
| replace focal loss with difficulty-aware attention (ablation) | enUNet · Pelvic prostate apex | DSC (%): 86.83 -> 88.21 | Improved |
| replace hybrid loss with difficulty-aware attention (ablation) | enUNet · Pelvic prostate middle | DSC (%): 90.08 -> 90.90 | Improved |
| use hybrid Dice+CE loss instead of focal loss (ablation) | enUNet · Pelvic prostate apex | DSC (%): 86.83 -> 86.71 | No clear effect |
| add enhancements (residual, deep supervision, attention fusion) to UNet (ablation) | UNet · Pelvic MRI | DSC (%) | Improved |
| add difficulty-aware attention mechanism (ablation) | enUNet · MR Brain challenge | DSC (%): about 3.5% average DSC | Improved |
| proposed single model vs top ensemble methods | Proposed · PROMISE12 prostate challenge | overall score: 88.28(3.02) vs 88.98(3.41) | Hurt |
Closest papers by meaning
- Adversarial Confidence Learning for Medical Image Segmentation and Synthesis (IJCV 2020 · similarity 0.90)
- Quantization of Fully Convolutional Networks for Accurate Biomedical Image Segmentation (CVPR 2018 · similarity 0.87)
- A case for reframing automated medical image classification as segmentation (NeurIPS 2023 · similarity 0.87)
- Tyche: Stochastic In-Context Learning for Medical Image Segmentation (CVPR 2024 · similarity 0.85)
- E2ENet: Dynamic Sparse Feature Fusion for Accurate and Efficient 3D Medical Image Segmentation (NeurIPS 2024 · similarity 0.85)
- Training Like a Medical Resident: Context-Prior Learning Toward Universal Medical Image Segmentation (CVPR 2024 · 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.
