Research map / Image generation / Editing and style transfer
Spectral Consistent Flow for One-step 3D Medical Image Translation
Haoqing Li, Jun Shi, Mingchao Li, Zehua Zhu, Qiwei Jia, Jiong SHI, Hong An
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 |
|---|---|---|---|
| replace deterministic MeanFlow with SC-Flow (Brownian bridge plus spectral corrector) | MeanFlow · MRI to Abeta-PET (ADNI) | PSNR: 22.16 -> 24.72 | Improved |
| replace deterministic MeanFlow with SC-Flow | MeanFlow · MRI to FDG-PET (NACC) | PSNR: 21.96 -> 27.41 | Improved |
| one-step SC-Flow instead of 100-NFE multi-step models | SC-Flow · MRI to Tau-PET (A4) | MAE: PASTA 2.84 -> 2.38 | Improved |
| use SC-Flow synthetic images for downstream AD classification | ResNet50 · NACC-FDG diagnosis | F1: 60.32 -> 71.77 | Improved |
| use SC-Flow synthetic images for downstream AD classification | SeResNext · A4-Tau diagnosis | AUC: 46.15 -> 70.19 | Improved |
| SC-Flow versus Pix2Pix on NACC-FDG AUC | SC-Flow · NACC-FDG diagnosis | AUC: 82.29 -> 81.25 | Hurt |
| use 24 radial bins instead of very fine bins (96) or too coarse (ablation) | Spectral Consistency Corrector · ablation (numbers not in extract) | accuracy | Improved |
| add spectral consistency corrector and Brownian bridge diffusion term (ablation) | SC-Flow · ablation (numbers not in extract) | translation accuracy | Improved |
Closest papers by meaning
- MAISI-v2: Accelerated 3D High-Resolution Medical Image Synthesis with Rectified Flow and Region-specific Contrastive Loss (AAAI 2026 · similarity 0.88)
- Knowledge Transfer Scaling Laws for 3D Medical Imaging (NeurIPS 2026 · similarity 0.85)
- STREAM: Stochastic Riemannian Flow Matching with Anisotropic Decoder for Digital Histopathology Image Generation (NeurIPS 2026 · similarity 0.84)
- MedFlowSeg: Flow Matching for Medical Image Segmentation with Frequency-Aware Attention (NeurIPS 2026 · similarity 0.84)
- Improving Progressive Generation with Decomposable Flow Matching (NeurIPS 2025 · similarity 0.84)
- VCT: Training Consistency Models with Variational Noise Coupling (ICML 2025 · similarity 0.84)
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.
