Research map / 3D vision / 3D generation
Sketch2CT: Multimodal Diffusion for Structure-Aware 3D Medical Volume Generation
Delin An, Chaoli Wang
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 TSFE text-guided sketch feature extractor (ablation) | Sketch2CT · CHAOS liver (CT) downstream segmentation | Dice: 0.864 -> 0.893 | Improved |
| add TSFE text-guided sketch feature extractor (ablation) | Sketch2CT · Decathlon heart (MRI) downstream segmentation | Dice: 0.683 -> 0.711 | Improved |
| add CGFM cross-modal global fusion (ablation) | Sketch2CT · CHAOS liver (CT) downstream segmentation | Dice: 0.825 -> 0.893 | Improved |
| add CGFM cross-modal global fusion (ablation) | Sketch2CT · Decathlon heart (MRI) downstream segmentation | Dice: 0.671 -> 0.711 | Improved |
| add segmentation latent diffusion model for mask generation (ablation) | Sketch2CT · CHAOS liver (CT) downstream segmentation | Dice: 0.642 -> 0.893 | Improved |
| add segmentation latent diffusion model for mask generation (ablation) | Sketch2CT · Decathlon heart (MRI) downstream segmentation | Dice: 0.545 -> 0.711 | Improved |
| combine sketch and text conditioning instead of sketch-only or text-only (ablation) | Sketch2CT · 3D mask generation (qualitative) | mask quality | Improved |
| increase sketch granularity from coarse to fine (ablation) | Sketch2CT · mask generation | Dice difference: 0.06 ± 0.02 | No clear effect |
| use Sketch2CT instead of Seg-Diff | Sketch2CT · Decathlon liver (CT) FID | FID: 34.8 -> 36.5 | Hurt |
| use Sketch2CT instead of Seg-Diff | Sketch2CT · CHAOS liver (CT) FID | FID: 37.8 -> 33.7 | Improved |
| use Sketch2CT instead of MedGen3D | Sketch2CT · Decathlon heart (MRI) FID | FID: 96.8 -> 65.1 | Improved |
| use Sketch2CT instead of Seg-Diff | Sketch2CT · Decathlon liver (CT) downstream segmentation | Dice: 0.887 -> 0.904 | Improved |
Closest papers by meaning
- Better Tokens for Better 3D: Advancing Vision-Language Modeling in 3D Medical Imaging (NeurIPS 2025 · similarity 0.88)
- UniPro: Unified Multi-Mode Medical Image Segmentation from 2D Images to 3D Volumes via Propagation (NeurIPS 2026 · similarity 0.87)
- LiFT: Lifted Inter-slice Feature Trajectories for 3D Image Generation from 2D Generators (NeurIPS 2026 · similarity 0.86)
- 3DMedAgent: Unified Perception-to-Understanding for 3D Medical Analysis (ICML 2026 · similarity 0.86)
- Knowledge Transfer Scaling Laws for 3D Medical Imaging (NeurIPS 2026 · similarity 0.86)
- CAT: Coordinating Anatomical-Textual Prompts for Multi-Organ and Tumor 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.
