Research map / Detection and segmentation / Open-set and few-shot detection
Bidirectional RNN-based Few Shot Learning for 3D Medical Image Segmentation
Soopil Kim, Sion An, Philip Chikontwe, Sang Hyun Park
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 a bidirectional GRU over adjacent-slice features (ablation) | FSS_base · BCV, spleen, 1-shot | Dice: 0.722±0.010 -> 0.816±0.028 | Improved |
| Fine-tune on target support data (ablation) | FSS_BiGRU, 5-shot · BCV, mean | Dice: 0.835 -> 0.866 | Improved |
| Fine-tune on target support data (ablation) | FSS_BiGRU, 5-shot · DECATHLON, spleen and liver | Dice: +3% | Improved |
| Fine-tune on target support data (ablation) | FSS_BiGRU, 5-shot · CTORG, kidney and liver | Dice: 0.716±0.03 -> 0.701±0.025; 0.788±0.003 -> 0.778±0.009 | Hurt |
| Fine-tune on target support data (ablation) | FSS_BiGRU, 5-shot · CTORG, bladder | Dice: 0.729±0.005 -> 0.731±0.001 | No clear effect |
| Increase support examples from 1-shot to 5-shot (ablation) | FSS_BiGRU · BCV, mean | Dice: 0.778 -> 0.835 | Improved |
| Use bidirectional recurrent slice context rather than independent slice predictions | Few-shot 3D segmentation · Qualitative coronal and sagittal views | segmentation smoothness | Improved |
Closest papers by meaning
- Combining Fully Convolutional and Recurrent Neural Networks for 3D Biomedical Image Segmentation (NeurIPS 2016 · similarity 0.89)
- Recurrent Mask Refinement for Few-Shot Medical Image Segmentation (ICCV 2021 · similarity 0.88)
- Upping the Game: How 2D U-Net Skip Connections Flip 3D Segmentation (NeurIPS 2024 · similarity 0.88)
- DoDNet: Learning To Segment Multi-Organ and Tumors From Multiple Partially Labeled Datasets (CVPR 2021 · similarity 0.88)
- Revisiting MAE Pre-training for 3D Medical Image Segmentation (CVPR 2025 · similarity 0.86)
- Integrating Deep Metric Learning with Coreset for Active Learning in 3D Segmentation (NeurIPS 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.
