Research map / Image generation / Fast diffusion sampling
SD-FSMIS: Adapting Stable Diffusion for Few-Shot Medical Image Segmentation
Meihua Li, Yang Zhang, Weizhao He, Hu Qu, Yisong Li
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 VTCT on top of the SII baseline (ablation) | SD-FSMIS · Abd-CT (Setting 1) | mean Dice: 80.11 -> 83.17 | Improved |
| add QE on top of the SII baseline (ablation) | SD-FSMIS · Abd-CT (Setting 1) | mean Dice: 80.11 -> 82.27 | Improved |
| combine QE and VTCT with SII (ablation) | SD-FSMIS · Abd-CT (Setting 1) | mean Dice: 83.17 -> 83.66 | Improved |
| use SD 1.5 instead of SD 2.1 backbone (ablation) | SD-FSMIS · Abd-CT (Setting 1) | mean Dice: 82.84 -> 83.66 | Improved |
| SD-FSMIS vs best prior method in-domain | SD-FSMIS · Abd-MRI (Setting 1) | mean Dice: 84.12 (DIFD) vs 83.16 | No clear effect |
| adapt pretrained SD with SQI and VTCT versus re-implemented DiffewS | SD-FSMIS · Abd-CT (Setting 2) | mean Dice: 74.93 -> 83.25 | Improved |
| use SD-based method in cross-domain transfer | SD-FSMIS · Cross-domain Abd-CT to MRI (Setting 1) | mean Dice: 71.14 (DIFD) -> 81.42 | Improved |
Closest papers by meaning
- FAMNet: Frequency-aware Matching Network for Cross-domain Few-shot Medical Image Segmentation (AAAI 2025 · similarity 0.89)
- VGD: Value-Guided Diffusion Toward High-Utility Medical Image Segmentation (AAAI 2026 · similarity 0.89)
- MedSegDiff-V2: Diffusion-Based Medical Image Segmentation with Transformer (AAAI 2024 · similarity 0.87)
- LLaDA-MedV: Exploring Large Language Diffusion Models for Biomedical Image Understanding (CVPR 2026 · similarity 0.87)
- Unleashing the Potential of the Diffusion Model in Few-shot Semantic Segmentation (NeurIPS 2024 · similarity 0.87)
- MedFlowSeg: Flow Matching for Medical Image Segmentation with Frequency-Aware Attention (NeurIPS 2026 · 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.
