Research map / Image generation / Text-to-image diffusion
Flexible Control of 3D CT Generation via Text and Semantically-Defined Segmentation Prompts
Weicheng Dai, Chenyu Wang, Shantanu Ghosh, Kayhan Batmanghelich
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 |
|---|---|---|---|
| use MedSyn2 instead of Text2CT | MedSyn2 · CT-Rate lung window, reports only | mean FID: 12.58 -> 9.59 | Improved |
| use MedSyn2 instead of Text2CT | MedSyn2 · CT-Rate bone window, reports only | mean FID: 7.36 -> 4.58 | Improved |
| use MedSyn2 instead of GenerateCT on axial plane | MedSyn2 · CT-Rate lung window XY | FID: 7.03 -> 10.96 | Hurt |
| use MedSyn2 instead of Text2CT | MedSyn2 · Pillar-0 T2I CLIP-Score | CLIP-Score (%): 25.80 -> 27.61 | Improved |
| use MedSyn2 instead of MedSyn | MedSyn2 · Zero-shot CT-CLIP classification | F1: 40.12 -> 48.40 | Improved |
| use MedSyn2 instead of MedSyn | MedSyn2 · Zero-shot CT-CLIP classification | Accuracy: 79.65 -> 63.55 | Hurt |
| add synthesized images to 50 real images for GGO segmentation | vanilla UNet · SemiSeg GGO | Dice (%): 58.78 -> 68.13 | Improved |
| train on synthesized images only instead of real | vanilla UNet · SemiSeg GGO | Dice (%): 58.78 -> 48.82 | Hurt |
| add convolution layer after each DiT block (ablation) | MedSyn2 · CT generation (qualitative) | visual artifacts | Improved |
| add gated attention to MHSA and MHCA (ablation) | MedSyn2 · CT generation (attention maps) | attention on abnormality terms | Improved |
| train without nodule labels (zero-shot nodule generation) (ablation) | MedSyn2 · Nodule synthesis (qualitative) | nodule quality | Mixed |
Closest papers by meaning
- GenerateCT: Text-Conditional Generation of 3D Chest CT Volumes (ECCV 2024 · similarity 0.89)
- CAT: Coordinating Anatomical-Textual Prompts for Multi-Organ and Tumor Segmentation (NeurIPS 2024 · similarity 0.89)
- Better Tokens for Better 3D: Advancing Vision-Language Modeling in 3D Medical Imaging (NeurIPS 2025 · similarity 0.87)
- Towards a Comprehensive, Efficient and Promptable Anatomic Structure Segmentation Model Using 3D Whole-Body CT Scans (AAAI 2025 · similarity 0.86)
- AHIVE: Anatomy-aware Hierarchical Vision Encoding for Interactive Radiology Report Retrieval (CVPR 2024 · similarity 0.86)
- SegVol: Universal and Interactive Volumetric Medical Image 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.
