🏥 Awesome Generative AI in Medical Imaging
原文可见Github仓库：https://github.com/Joker-ZXR/Awesome-Generative-AI-in-Medical-Imaging
🌟 Highlights
A curated collection of datasets, methods, and resources for Generative AI in Medical Imaging
Compiled from "Generative Artificial Intelligence in Medical Imaging: Foundations, Progress, and Clinical Translation"
🎯 What Makes This Different:
- ✨ Complete Clinical Coverage: First survey mapping generative AI across diagnosis, treatment, and prognosis
- 🎓 Standardized Evaluation: Three-tiered framework (image fidelity, feature consistency, clinical relevance)
- 🔗 Ready-to-Use Resources: All datasets and methods with direct links and implementation details
- 📊 Comprehensive Scope: 95 datasets, 175 methods, 10+ imaging modalities
📑 Table of Contents
📚 Quick Navigation
- [🌟 Highlights](#🌟 Highlights)
- [🚀 Quick Start](#🚀 Quick Start)
- [📊 Datasets](#📊 Datasets)
- [🌐 Whole Body](#🌐 Whole Body)
- [🧠 Head and Neck](#🧠 Head and Neck)
- [🔘 Chest](#🔘 Chest)
- [❤️ Cardiac](#❤️ Cardiac)
- [🔹 Abdomen](#🔹 Abdomen)
- [🦴 Musculoskeletal](#🦴 Musculoskeletal)
- [🔗 Multimodal](#🔗 Multimodal)
- [🔬 Methods](#🔬 Methods)
- [🧹 Denoising & Artifact Removal](#🧹 Denoising & Artifact Removal)
- [🔄 Image Reconstruction](#🔄 Image Reconstruction)
- [🔍 Super-Resolution](#🔍 Super-Resolution)
- [🎨 Unconditional Synthesis](#🎨 Unconditional Synthesis)
- [🎯 Conditional Synthesis](#🎯 Conditional Synthesis)
- [💊 Treatment Planning](#💊 Treatment Planning)
- [📈 Disease Progression](#📈 Disease Progression)
- [🏗️ Foundation Models](#🏗️ Foundation Models)
- [🤝 Contributing](#🤝 Contributing)
- [🙏 Acknowledgments](#🙏 Acknowledgments)
- [📜 License](#📜 License)
- [📖 Citation](#📖 Citation)
🚀 Quick Start
🎯 What You'll Find
| Resource | Description | Count |
|---|---|---|
| 📊Public Datasets | Curated medical imaging datasets with direct links | 95 |
| 🔬Research Methods | State-of-the-art generative AI methods | 175 |
| 💻Code Repositories | GitHub/project links for implementations | 90+ |
| 📝Loss Functions | Detailed training objectives for each method | All |
| 🏥Clinical Applications | Diagnosis, treatment, and prognosis workflows | Full Coverage |
📊 Datasets
💡 Tip: All 95 datasets include direct download links. Click on dataset names to access them immediately.
📍 Browse by Anatomy
| Anatomical Region | Datasets | Modalities | Jump |
|---|---|---|---|
| 🌐Whole Body | 14 | CT, PET-CT, MRI, Pathology | [View](#Anatomical Region Datasets Modalities Jump 🌐Whole Body 14 CT, PET-CT, MRI, Pathology View 🧠Head & Neck 23 CT, PET-CT, MRI, US, OCT, Fundus, Pathology View 🔘Chest 16 X-ray, CT, MRI, US, Pathology View ❤️Cardiac 12 MRI, Ultrasound View 🔹Abdomen 15 CT, MRI, US, Pathology View 🦴Musculoskeletal 5 MRI View 🔗Multimodal 10 Image-Text, Clinical Data View) |
| 🧠Head & Neck | 23 | CT, PET-CT, MRI, US, OCT, Fundus, Pathology | [View](#Anatomical Region Datasets Modalities Jump 🌐Whole Body 14 CT, PET-CT, MRI, Pathology View 🧠Head & Neck 23 CT, PET-CT, MRI, US, OCT, Fundus, Pathology View 🔘Chest 16 X-ray, CT, MRI, US, Pathology View ❤️Cardiac 12 MRI, Ultrasound View 🔹Abdomen 15 CT, MRI, US, Pathology View 🦴Musculoskeletal 5 MRI View 🔗Multimodal 10 Image-Text, Clinical Data View) |
| 🔘Chest | 16 | X-ray, CT, MRI, US, Pathology | [View](#Anatomical Region Datasets Modalities Jump 🌐Whole Body 14 CT, PET-CT, MRI, Pathology View 🧠Head & Neck 23 CT, PET-CT, MRI, US, OCT, Fundus, Pathology View 🔘Chest 16 X-ray, CT, MRI, US, Pathology View ❤️Cardiac 12 MRI, Ultrasound View 🔹Abdomen 15 CT, MRI, US, Pathology View 🦴Musculoskeletal 5 MRI View 🔗Multimodal 10 Image-Text, Clinical Data View) |
| ❤️Cardiac | 12 | MRI, Ultrasound | [View](#Anatomical Region Datasets Modalities Jump 🌐Whole Body 14 CT, PET-CT, MRI, Pathology View 🧠Head & Neck 23 CT, PET-CT, MRI, US, OCT, Fundus, Pathology View 🔘Chest 16 X-ray, CT, MRI, US, Pathology View ❤️Cardiac 12 MRI, Ultrasound View 🔹Abdomen 15 CT, MRI, US, Pathology View 🦴Musculoskeletal 5 MRI View 🔗Multimodal 10 Image-Text, Clinical Data View) |
| 🔹Abdomen | 15 | CT, MRI, US, Pathology | [View](#Anatomical Region Datasets Modalities Jump 🌐Whole Body 14 CT, PET-CT, MRI, Pathology View 🧠Head & Neck 23 CT, PET-CT, MRI, US, OCT, Fundus, Pathology View 🔘Chest 16 X-ray, CT, MRI, US, Pathology View ❤️Cardiac 12 MRI, Ultrasound View 🔹Abdomen 15 CT, MRI, US, Pathology View 🦴Musculoskeletal 5 MRI View 🔗Multimodal 10 Image-Text, Clinical Data View) |
| 🦴Musculoskeletal | 5 | MRI | [View](#Anatomical Region Datasets Modalities Jump 🌐Whole Body 14 CT, PET-CT, MRI, Pathology View 🧠Head & Neck 23 CT, PET-CT, MRI, US, OCT, Fundus, Pathology View 🔘Chest 16 X-ray, CT, MRI, US, Pathology View ❤️Cardiac 12 MRI, Ultrasound View 🔹Abdomen 15 CT, MRI, US, Pathology View 🦴Musculoskeletal 5 MRI View 🔗Multimodal 10 Image-Text, Clinical Data View) |
| 🔗Multimodal | 10 | Image-Text, Clinical Data | [View](#Anatomical Region Datasets Modalities Jump 🌐Whole Body 14 CT, PET-CT, MRI, Pathology View 🧠Head & Neck 23 CT, PET-CT, MRI, US, OCT, Fundus, Pathology View 🔘Chest 16 X-ray, CT, MRI, US, Pathology View ❤️Cardiac 12 MRI, Ultrasound View 🔹Abdomen 15 CT, MRI, US, Pathology View 🦴Musculoskeletal 5 MRI View 🔗Multimodal 10 Image-Text, Clinical Data View) |
🌐 Whole Body
View 14 Whole Body Datasets
Comprehensive datasets covering multiple organs and body regions
| Dataset | Modality | Scale | Source |
|---|---|---|---|
| AutoPET | 3D PET-CT | 1014 volumes | Grand Challenge |
| AutoPETIII | 3D PET-CT | 1614 volumes | Grand Challenge |
| CPIA | Pathology | 21.4M WSI | GitHub |
| CT-ORG | 3D CT | 140 volumes | TCIA |
| DeepLesion | 2D CT | 32.7K images | nihcc.app.box.com |
| FLARE24 Task1 | 3D CT | 10K volumes | Challenge Platform |
| MedMNIST | 2D & 3D | 708K 2D images & 10K 3D volumes | Zenodo/DOI |
| NAFLD | Pathology | 119.8K WSI | osf.io |
| PreCT-160K | 3D CT | 160K volumes | Hugging Face |
| TotalSegmentator | 3D CT | 1204 volumes | Zenodo/DOI |
| TotalSegmentator MRI | 3D MRI | 298 volumes | Zenodo/DOI |
| TotalSegmentator MRI v2 | 3D MRI | 616 volumes | Zenodo/DOI |
| TotalSegmentator v2 | 3D CT | 1228 volumes | Zenodo/DOI |
| ULS | 3D CT | 38.8K volumes | Grand Challenge |
🧠 Head and Neck
View 23 Head & Neck Datasets
Datasets for brain, thyroid, retina, and head/neck imaging
| Dataset | Modality | Scale | Source |
|---|---|---|---|
| AIROGS | Fundus | 101.4K images | Grand Challenge |
| AOMIC | 3D MRI | 1370 volumes | nilab-uva.github.io |
| BraTS2023-MEN | 3D MRI | 1650 volumes | Synapse |
| BraTS21 | 3D MRI | 2040 volumes | Synapse |
| CrossMoDA2021 | 3D MRI | 349 volumes | Grand Challenge |
| CrossMoDA2023 | 3D MRI | 983 volumes | Synapse |
| Diabetic Retinopathy Arranged | Fundus | 35.1K images | tianchi.aliyun.com |
| Diff5T | 3D MRI | 14.65K volumes | ScienceDB |
| HECKTOR2022 | 3D PET-CT | 882 volumes | Grand Challenge |
| INSTANCE2022 | 3D CT | 200 volumes | Grand Challenge |
| IXI Dataset | 3D MRI | 600 volumes | brain-development.org |
| LAG | Fundus | 11.7K images | GitHub |
| OCT2017 | OCT | 35.1K images | data.mendeley.com |
| ODIR-5K | Fundus | 5000 images | Grand Challenge |
| OpticNerveSeg | 3D MRI | 151 volumes | ScienceDB |
| OSCC | Pathology | 1224 WSI | data.mendeley.com |
| PatchCamelyon | Pathology | 327.7K WSI | GitHub |
| Retinal OCT-C8 | OCT | 24K images | Kaggle |
| SegRap2023 | 3D CT | 200 volumes | Grand Challenge |
| TN-SCUI2020 | 2D US | 4554 images | Grand Challenge |
| TN3K | 2D US | 3493 images | GitHub |
| Ultrasound Nerve Segmentation | 2D US | 11.1K images | Kaggle |
| fastMRI_Brain | 2D MRI | 6970 images | Project Site |
🔘 Chest
View 16 Chest Datasets
Datasets for lung, breast, and thoracic imaging
| Dataset | Modality | Scale | Source |
|---|---|---|---|
| ACRIN-Contralateral-Breast-MR | 3D MRI | 984 volumes | TCIA |
| ATM22 | 3D CT | 500 volumes | Grand Challenge |
| BRAX | 2D X-ray | 40.9K images | PhysioNet |
| BUSI | US | 780 images | scholar.cu.edu.eg |
| Breakhis | Pathology | 7909 WSI | opendatalab.com |
| CheXchoNet | 2D X-ray | 71.6K images | PhysioNet |
| ISPY1-Tumor-SEG-Radiomics | 3D MRI | 483 volumes | TCIA |
| LIDC-IDRI | 2D CT | 1010 images | TCIA |
| LNQ2023 | 3D CT | 513 volumes | Grand Challenge |
| LUNA16 | 3D CT | 888 volumes | Grand Challenge |
| MIST-HER2 | Pathology | 22.7K WSI | drive.google.com |
| SARS-COV-2 Ct-Scan | 2D CT | 2482 images | Kaggle |
| SIIM-FISABIO-RSNA COVID-19 | 2D X-ray | 7597 images | Kaggle |
| TDSC-ABUS2023 | US | 200 volumes | Grand Challenge |
| WSSS4LUAD | Pathology | 10K WSI | Grand Challenge |
| fastMRI_Breast | 2D MRI | 300 images | Project Site |
❤️ Cardiac
View 12 Cardiac Datasets
Datasets for heart structure and function imaging
| Dataset | Modality | Scale | Source |
|---|---|---|---|
| ACDC | 3D MRI | 150 volumes | humanheart-project.creatis.insa-lyon.fr |
| CMRxRecon | 2Dt MRI | 300 volumes | Synapse |
| Cardiac MRI Dataset | 2Dt MRI | 7980 volumes | data.nvision.eecs.yorku.ca |
| Cardiac super-resolution label maps | 2Dt MRI | 1331 volumes | data.mendeley.com |
| EchoNet-Dynamic | US | 10K videos | echonet.github.io |
| EchoNet-LVH | US | 12K videos | echonet.github.io |
| GANcMRI | US | 45.5K videos | biobank.ndph.ox.ac.uk |
| Harvard Cardiac MR Center Dataverse | 2Dt MRI | 108 volumes | Harvard Dataverse |
| M&Ms Challenge | 3D MRI | 375 volumes | mega.nz |
| M&Ms-2 Challenge | 3D MRI | 360 volumes | mega.nz |
| MICCAI 2024 CARE LAScarQS++ | 3D MRI | 194 volumes | zmic.org.cn |
| OCMR | 2Dt MRI | 165 volumes | ocmr.info |
🔹 Abdomen
View 15 Abdomen Datasets
Datasets for liver, kidney, pancreas, and GI tract imaging
| Dataset | Modality | Scale | Source |
|---|---|---|---|
| AbdomenAtlas 1.0 Mini | 3D CT | 5195 volumes | Hugging Face |
| AbdomenCT-1K | 3D CT | 1112 volumes | GitHub |
| EndoSLAM | US | 1020 videos | GitHub |
| FLARE 2024 Task3 | 3D MRI | 4817 volumes | Challenge Platform |
| FLARE2022 | 3D CT | 2300 volumes | Grand Challenge |
| FLARE2023 | 3D CT | 4500 volumes | Challenge Platform |
| GasHisSDB | Pathology | 245.2K images | gitee.com |
| ISBI 2025 FUGC | US | 890 images | Zenodo/DOI |
| LIMUC | US | 1043 videos | Zenodo/DOI |
| NCT-CRC-HE | Pathology | 100K WSI | Zenodo/DOI |
| PI-CAI | 3D MRI | 1500 volumes | Zenodo/DOI |
| RenalCell | Pathology | 625.1K WSI | Zenodo/DOI |
| SUN | US | 1018 videos | amed8k.sundatabase.org |
| SegPANDA200 | Pathology | 100.9K images | drive.google.com |
| UW-Madison GI Tract Image | 2D MRI | 38.5K images | Kaggle |
🦴 Musculoskeletal and Other
View 5 Musculoskeletal Datasets
Datasets for bone, joint, and spine imaging
| Dataset | Modality | Scale | Source |
|---|---|---|---|
| MRNet (Knee) | 3D MRI | 1370 volumes | Project Site |
| SKM-TEA (Knee) | 3D MRI | 155 volumes | Project Site |
| SPIDER (Spine) | 3D MRI | 257 volumes | Zenodo/DOI |
| Wrist Dataset | 3Dt MRI | 55 volumes | data.mendeley.com |
| fastMRI_knee | 2D MRI | 1398 images | Project Site |
🔗 Multimodal Datasets
View 10 Multimodal Datasets
Datasets combining images with text, clinical data, or genomics
| Dataset | Modality | Scale | Source |
|---|---|---|---|
| CheXpertPlus | X-ray-Text | 223K images, 223K texts | Project Site |
| Duke Breast Cancer MRI | Genomic&MRI-Clinical data | 922 cases | sites.duke.edu |
| I-SPY2 | MRI-Clinical data | 719 cases | TCIA |
| MedICaT | Multimodal Images-Text | 217K images, 217K texts | GitHub |
| Medical-CXR-VQA | X-ray-Text | 377K images, 780K texts | GitHub |
| Medtrinity-25M | Multimodal Images-Text | 25M images, 25M texts | GitHub |
| OpenPath | Pathology-Text | 208K images, 208K texts | Hugging Face |
| PMC-OA | Multimodal Images-Text | 1.6M images, 1.6M texts | Hugging Face |
| PadChest | X-ray-Text | 160K images, 109K texts | bimcv.cipf.es |
| Quilt-1M | Pathology-Text | 1M images, 1M texts | quilt1m.github.io |
🔬 Methods
💡 Tip: 90+ methods include direct links to code repositories and project pages.
🎯 Browse by Application
| Application Category | Methods | Models | Jump |
|---|---|---|---|
| 🧹Denoising & Artifact Removal | 16 | GAN, CNN, Diffusion, Mamba, Transformer | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
| 🔄Image Reconstruction | 47 | GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
| 🔍Super-Resolution | 18 | GAN, CNN, Diffusion, Transformer, Mamba | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
| 🎨Unconditional Synthesis | 11 | GAN, VAE, Diffusion, Mamba | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
| 🎯Conditional Synthesis | 38 | GAN, VAE, Diffusion, Transformer, Mamba | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
| 💊Treatment Planning | 23 | GAN, CNN, Diffusion, VAE, Transformer, Mamba | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
| 📈Disease Progression | 12 | GAN, CNN, Diffusion, VAE, AR | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
| 🏗️Foundation Models | 10 | Diffusion, CNN, Transformer, VAE | [View](#Application Category Methods Models Jump 🧹Denoising & Artifact Removal 16 GAN, CNN, Diffusion, Mamba, Transformer View 🔄Image Reconstruction 47 GAN, CNN, Diffusion, VAE, Transformer, Mamba, AR View 🔍Super-Resolution 18 GAN, CNN, Diffusion, Transformer, Mamba View 🎨Unconditional Synthesis 11 GAN, VAE, Diffusion, Mamba View 🎯Conditional Synthesis 38 GAN, VAE, Diffusion, Transformer, Mamba View 💊Treatment Planning 23 GAN, CNN, Diffusion, VAE, Transformer, Mamba View 📈Disease Progression 12 GAN, CNN, Diffusion, VAE, AR View 🏗️Foundation Models 10 Diffusion, CNN, Transformer, VAE View) |
🧹 Denoising and Artifact Removal
View 16 Denoising & Artifact Removal Methods
Methods for noise reduction and artifact suppression in medical images
| Method / Publication | Model | Application | Loss Functions |
|---|---|---|---|
| PWGAN-WSHL (2021) | GAN | Low-dose CT denoising | WGAN loss, L1 loss, MSE loss, structural loss |
| DenoMamba (2024) | Mamba | Low-dose CT denoising | L1 loss |
| m-WGAN (2019) | GAN | CT image artifact removal | WGAN loss, MSE loss |
| TT U-Net (2023) | Transformer | CT image artifact removal | L1 loss, adversarial loss |
| CoreDiff (2024) | Diffusion model | Low-dose CT denoising | Diffusion denoising loss |
| PFGM++ (2024) | Diffusion model | Photon-counting CT denoising | Diffusion denoising loss |
| Yang et al. (2021) | CNN | PET image denoising and artifact Removal | MSE loss |
| Hu et al. (2020) | GAN | PET image denoising and artifact Removal | WGAN loss, MSE loss, gradient difference loss, content loss, ssim loss |
| PT-WGAN (2020) | GAN | PET image denoising | Adversarial loss, MSE loss, ssim loss, perceptual loss |
| Gong et al. (2024) | Diffusion model | PET image denoising | Diffusion denoising loss |
| Yu et al. (2024) | Diffusion model | PET image denoising | Diffusion denoising loss |
| RED-WGAN (2019) | GAN | MRI image denoising | WGAN loss, MSE loss, perceptual loss, VGG loss |
| Chung et al. (2022) | Diffusion model | MRI image denoising | Diffusion denoising loss |
| CNCL (2022) | GAN | MR, CT and PET image denoising | Content loss, noise loss, GAN loss |
| Lim et al. (2023) | GAN | MRI image artifact removal | WGAN loss, L1 loss, VGG loss |
| PFAD (2024) | Diffusion model | MRI image artifact removal | Diffusion denoising loss |
🧩 Image Reconstruction
CT Reconstruction
View 12 CT Reconstruction Methods
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| DL-recon (2022) | GAN | CBCT-to-CT reconstruction | Adversarial loss, L1 loss |
| Pradhan et al. (2023) | GAN | 2D-to-3D CT reconstruction | L1 loss, BCE loss, adversarial loss |
| HyperNeRFGAN (2024) | GAN | X-ray-to-CT reconstruction | StyleGAN2Loss |
| Krishnan et al. (2024) | GAN | Low-dose CT reconstruction | L1 loss, adversarial loss, reconstruction loss |
| MambaMIR (2025) | GAN, Mamba | Low-dose CT/PET reconstruction | Adversarial loss, Charbonnier loss，image loss, frequency loss |
| SI-GAN (2019) | GAN | Limited-angle CT reconstruction | Adversarialloss, sinogram loss, reconstruction loss |
| DOLCE (2023) | Diffusion model | Limited-angle CT reconstruction | Diffusion denoising loss |
| Lopez-Montes et al. (2024) | Diffusion model | Limited-angle CT reconstruction | Diffusion denoising loss |
| TIFA (2024) | Diffusion model | Limited-angle CT reconstruction | Diffusion denoising loss |
| Xia et al. (2024) | Diffusion model | Sparse-view CT reconstruction | Diffusion denoising loss |
| CDDM (2024) | Diffusion model | Sparse-view CT reconstruction | Diffusion denoising loss |
| SWORD (2024) | Diffusion model | Sparse-view CT reconstruction | Diffusion denoising loss |
PET Reconstruction
View 12 PET Reconstruction Methods
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| NADRU (2020) | CNN | Low dose PET reconstruction | Dice loss, BCE loss, general and adaptive robust loss, ssim loss |
| Shi et al. (2023) | CNN | Low dose PET reconstruction | L1 loss, image domain loss, gradient difference loss, LIP loss |
| CPR-CNN (2024) | CNN | Low dose PET reconstruction | Reconstruction loss, cycle consistency loss |
| DGLM (2024) | CNN | Low count PET reconstruction | MSE loss, ssim loss |
| Lei et al. (2019) | GAN | Low count PET reconstruction | Cycle-consistent adversarial loss, gradient descent loss, mean p-norm distance loss |
| Task-GAN (2019) | GAN | Ultra-low dose PET reconstruction | L1 loss, adversarial loss, regression loss |
| AR-GAN (2022) | GAN | Low dose PET reconstruction | L1 loss, adversarial loss, cross-entropy loss |
| DDPET-3D (2024) | Diffusion model | Low dose PET reconstruction | Diffusion denoising loss |
| Wikberg et al. (2024) | CNN | Sparsely acquired projections PET reconstruction | L1 loss, MSE loss |
| MMJSD (2024) | VAE | Bimodal PET/MRI reconstruction | Negative log-likelihood loss, KL loss |
| Singh et al. (2024) | Diffusion model | 2D/3D PET reconstruction | Poisson Log-Likelihood loss, Diffusion denoising loss |
| MC-Diffusion (2024) | Diffusion model | PET-MRI reconstruction | Diffusion denoising loss |
MRI reconstruction
Show 16 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| Wang et al. (2019) | GAN | MRI reconstruction | Content loss, perceptual loss, adversarial loss, dc loss |
| rsGAN (2020) | GAN | Multi-contrast MRI reconstruction | L1 loss, perceptual loss, adversarial loss, dc loss |
| Kelkar et al. (2021) | GAN | MRI reconstruction | MSE loss, log-likelihood loss, TV loss, dc loss |
| SwinMR (2022) | Transformer | MRI reconstruction | Pixel-wise Charbonnier loss, frequency Charbonnier loss |
| KM-MAML (2023) | CNN | MRI reconstruction | L1 reconstruction loss, dc loss |
| MambaMIR (2024) | Mamba | MRI reconstruction | Adversarial loss, image loss, kspace loss, perceptual loss, dc loss |
| DM-Mamba (2025) | Mamba | MRI reconstruction | L1 loss, dc loss |
| MambaRoll (2024) | Mamba, AR | MRI reconstruction | Kspace loss, cascade loss, dc loss |
| HFS-SDE (2024) | Diffusion model | MRI reconstruction | Diffusion denoising loss, dc loss |
| JSMoCo (2025) | Diffusion model | MRI reconstruction | Diffusion denoising loss, dc loss |
| AID (2025) | Diffusion model, AR | MRI reconstruction | Diffusion denoising loss, dc loss |
| Kofler et al. (2020) | CNN | Cardiac cine MRI reconstruction | L2 loss, dc loss |
| Qiu et al. (2024) | Diffusion model | Cardiac cine MRI reconstruction | Diffusion denoising loss, dc loss |
| DiffCMR (2024) | Diffusion model | Cardiac cine MRI reconstruction | Diffusion denoising loss, dc loss |
| FedGIMP (2023) | GAN | Federated MRI reconstruction | Logistic adversarial loss, local reconstruction loss, dc loss |
| FedGAT (2025) | VAE, Transformer, AR | Federated MRI reconstruction | Perceptual loss, adversarial loss, cross-entropy loss, MSE loss, dc loss |
Other Modalities
Show 7 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| DDRM (2023) | Diffusion model | US image reconstruction | Diffusion denoising loss |
| DRUSvar (2024) | Diffusion model | US image reconstruction | Diffusion denoising loss |
| Lan et al. (2023) | Diffusion model, GAN | US image reconstruction | Diffusion denoising loss |
| Merino et al. (2024) | Diffusion model, GAN | US image reconstruction | Adversarial loss, Diffusion denoising loss |
| DM-RE2I (2023) | Diffusion model | EEG to image reconstruction | Diffusion denoising loss |
| PAT-Diffusion (2023) | Diffusion model | Photoacoustic tomography reconstruction | Diffusion denoising loss |
| Tong et al. (2023) | Diffusion model | Photoacoustic tomography reconstruction | Diffusion denoising loss |
🔎 Super-Resolution
Temporal super-resolution
Show 9 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| Ren et al. (2021) | CNN | Video super‐resolution | L1 loss |
| Song (2022) | CNN, Transformer | Video super‐resolution | MSE loss |
| VSRResFeatGAN (2019) | GAN | Video super‐resolution | Adversarial loss, perceptual loss, charbonnier loss, |
| MFIN (2019) | CNN | 4D MRI Temporal super‐resolution | Cycle consistency loss, recon loss, ssim loss, |
| SVIN (2020) | CNN | 4D MRI Temporal super‐resolution | Similarity loss, smoothness regularization loss, regression loss |
| DDoS-Unet (2024) | CNN | 4D MRI Temporal super‐resolution | Perceptual loss, L1 loss |
| MPVF (2023) | CNN, Transformer | 4D MRI Temporal super‐resolution | Charbonnier loss |
| UVI-Net (2024) | CNN, Transformer | 4D MRI Temporal super‐resolution | NCC loss, gradient loss |
| DDM (2022) | Diffusion model | 4D MRI Temporal super‐resolution | Diffusion denoising loss, NCC loss, KL loss |
Spatial super-resolution
Show 9 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| GAN-CIRCLE (2019) | GAN | CT super‐resolution | Adversarial loss, cycle-consistency loss, identity loss, joint sparsifying transform loss |
| TTSR-FD (2021) | GAN | X-ray super‐resolution | Frequency domain loss, perpetual loss, adversarial loss, |
| SOUP-GAN (2022) | GAN | MRI super‐resolution | Adversarial loss, perceptual loss |
| DWT-SRGAN (2022) | GAN | MRI super‐resolution | Perpetual loss, adversarial loss, wavelet loss |
| HA-GAN (2022) | GAN | CT/MRI super‐resolution | GAN loss, reconstruction loss |
| Huang et al. (2024) | Transformer, CNN, | US/OCT/Endoscope/CT/MRI super‐resolution | Charbonnier loss, L1 loss |
| UHRCT_SR (2023) | Diffusion model | CT super‐resolution | Diffusion denoising loss |
| PartDiff (2023) | Diffusion model | MRI super‐resolution | Diffusion denoising loss |
| Deform-Mamba (2024) | Mamba | MRI super‐resolution | L1 loss, CE loss |
🎲 Unconditional Synthesis
Show 11 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| Danu et al. (2019) | VAE，GAN | Blood vessel surfaces synthesis | MSE loss，adversarial loss |
| Syn-Net (2020) | GAN | 2D brain MRI synthesis | L1 loss, perceptual loss, adversarial loss |
| Chong and Ho (2021) | GAN | 3D brain MRI synthesis | Adversarial loss，GAN loss |
| 3D-StyleGAN (2021) | GAN | 3D MRI synthesis | MSE loss, Logistic loss |
| Txurio et al. (2023) | Diffusion model | 2D CT synthesis | Diffusion denoising loss |
| MRGen (2024) | Diffusion model， VAE | 2D MRI synthesis | Diffusion denoising loss |
| VM-DDPM (2024) | Diffusion model, Mamba | 2D X-ray/MRI synthesis | Diffusion denoising loss, GAN loss, BCE Loss |
| GH-DDM (2023) | Diffusion model | 2D X-ray/CT/MRI/OCT synthesis | Diffusion denoising loss |
| Medicaldiffusion (2023) | Diffusion model | 3D CT/MRI synthesis | Diffusion denoising loss |
| DRDM (2024) | Diffusion model | 3D CT/MRI synthesis | Distance error loss, angle error loss, regularization loss |
| 3D MedDiffusion (2024) | Diffusion model | 3D CT/MRI synthesis | Vector quantization loss, adversarial loss, tri-plane loss, Diffusion denoising loss |
🎯 Conditional Synthesis
Text-to-Image Synthesis
Show 15 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| Campello et al. (2022) | GAN | Clinical information-to-MRI | Adversarial loss, cycle-consistency loss |
| CHeart (2023) | VAE | Clinical information-to-MRI | KL loss, log-likelihood loss |
| TUMSyn (2024) | Transformer, CNN | Clinical information-to-MRI | Contrastive loss, similarity loss |
| Del Castillo et al. (2025) | Diffusion model, VAE | Clinical information-to-MRI | Diffusion denoising loss |
| TaDiff (2025) | Diffusion model | Clinical information-to-MRI | Diffusion denoising loss，dice loss |
| MAISI (2024) | Diffusion model | Clinical information-to-CT | Diffusion denoising loss |
| EchoDiffusion (2023) | Diffusion model | Clinical information-to-video | Diffusion denoising loss |
| Kawata et al. (2024) | Diffusion model, VAE | Clinical information-to-chest CT synthesis | Diffusion denoising loss, similarity loss |
| GenerateCT (2024) | Diffusion model，Transformer | Report-to-chest CT synthesis | Diffusion denoising loss, perceptual loss, adversarial loss |
| MedSyn (2024) | Diffusion model, VAE | Report-to-chest CT synthesis | Diffusion denoising loss, KL loss |
| DCM-VLC (2024) | Diffusion model， GAN | Text-guided CT synthesis | Diffusion denoising loss，adversarial loss |
| MediSyn (2025) | Diffusion model, VAE | Text-guided diverse synthesis | Diffusion denoising loss |
| TextoMorph (2024) | Diffusion model | Text-guided tumor synthesis | Diffusion denoising loss, contrastive loss |
| Diff-CXR (2024) | Diffusion model，Transformer | Report-to-CXR synthesis | Diffusion denoising loss，InfoNCE loss, BCE loss |
| Chest-diffusion (2024) | Diffusion model, VAE | Report-to-CXR synthesis | Diffusion denoising loss， contrast loss |
Image-to-Image Synthesis
Show 17 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| Ben-Cohen (2019) | GAN | CT-to-PET translation | Adversarial loss, MSE loss, L1 loss |
| Jiao et al. (2020) | GAN | US-to-MRI translation | Latent space loss, appearance loss, structural consistency loss, adversarial loss |
| sc-cycleGAN (2020) | GAN | MR-to-CT translation | Adversarial loss, cycle-consistency loss, structure-consistency loss |
| Gong et al. (2020) | GAN | MRI-to-PET translation | Adversarial loss, cycle-consistency loss |
| GLFC (2025) | Mamba | CBCT-to-CT translation | Multiple contrast Loss |
| EGDiff (2024) | Diffusion model | CBCT-to-CT translation | Diffusion denoising loss, MSE loss |
| DiffMa (2024) | Diffusion model, Mamba | CT-to-MRI translation | Diffusion denoising loss, infoNCE loss |
| MIDiffusion (2024) | Diffusion model | MRI cross-modality translation | Mutual information diffusion denoising loss |
| Yan et al. (2022) | GAN | Multimodal MRI completion | Adversarial loss, cycle-consistency loss |
| Raad et al. (2024) | GAN | Multimodal MRI completion | Adversarial loss, MAE loss |
| CKG-GAN (2024) | GAN | Multimodal MRI completion | Cross-dimensional knowledge loss + adversarial |
| Zhang et al. (2024) | GAN | Multimodal MRI completion | Synthesis loss, reconstruction loss, adversarial loss, |
| AutoSyncoder (2020) | GAN，VAE | Multimodal MRI completion | Adversarial loss, negative log-likelihood loss |
| I2I-Mamba (2024) | Mamba | Multimodal MRI completion | Adversarial loss, pixel-wise loss |
| ResViT (2022) | Transformer | Multimodal MRI completion | L1 loss, adversarial loss |
| MMT (2023) | Transformer | Multimodal MRI completion | Synthesis loss, reconstruction loss, adversarial loss, |
| FgC2F-UDiff (2024) | Diffusion model | Multimodal MRI completion | Diffusion denoising loss |
Anatomically-Guided Image Synthesis
Show 6 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| CG-SAMR (2021) | GAN | Anatomy-guided CT synthesis | Adversarial loss, confidence map loss,feature matching loss, shape consistency loss |
| Shen et al. (2023) | GAN | Anatomy-guided CXR synthesis | Reconstruction loss, Perceptual loss, Adversarial loss |
| Hou et al. (2023) | GAN | Anatomy-guided fundus image synthesis | Wasserstein GAN loss, feature matching loss, KL-loss |
| Real-ESRGAN (2024) | GAN | Anatomy-guided fundus image synthesis | Adversarial loss, perceptual loss, L1 loss, L1_seg loss |
| LN-Gen (2024) | Diffusion model | Anatomy-guided rectal lymph nodes synthesis | Diffusion denoising loss, adapter loss |
| SegGuidedDiff (2024) | Diffusion model | Anatomy-guided MRI synthesis | Diffusion denoising loss |
💊 Treatment Planning and Dynamic Intervention
Generation for Treatment Planning
Show 9 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| DoseNet (2018) | CNN | Radiation dose prediction | L2 loss |
| C3D (2021) | CNN | Radiation dose prediction | L1 loss |
| Radonic et al. (2024) | CNN | Radiation dose prediction | MSE loss |
| TransDose (2023) | Transformer | Radiation dose prediction | Cross entropy loss，Charbonnier Loss |
| VQGAN_TATrans (2024) | GAN，VAE，Transformer | Brain tumor prediction | Pixel differences loss，perceptual loss，feature matching loss，gradient loss，codebook loss |
| PC-DDPM (2024) | Diffusion model | Real-time tumor tracking | Diffusion denoising loss，cycle-consistency loss |
| DiffDP (2023) | Diffusion model | Radiation dose prediction | Diffusion denoising loss |
| SP-DiffDose (2023) | Diffusion model，Transformer | Radiation dose prediction | Diffusion denoising loss |
| MD-Dose (2024) | Diffusion model，Mamba | Radiation dose prediction | Diffusion denoising loss |
Intraoperative navigation: Dynamic image synthesis
Show 14 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| SVIN (2020) | CNN | 4D dynamic MRI synthesis | Similarity loss, smoothness regularization loss, regression loss |
| DragNet (2023) | CNN | 2Dt cardiac MR synthesis | ELBO loss, KL loss, similarity loss |
| Quintero et al. (2024) | CNN | 4D dynamic MRI synthesis | RMSE loss |
| MPVF (2023) | CNN, Transformer | 4D dynamic MRI synthesis | Charbonnier loss |
| UVI-Net (2024) | CNN, Transformer | 4D dynamic MRI synthesis | NCC loss, gradient loss |
| TAV-GAN (2021) | GAN | 4D dynamic MRI synthesis | Temporally aware loss, SSIM loss, L1 loss |
| Thummerer et al. (2022) | GAN | 4D CT synthesis | MSE loss |
| REGAIN (2023) | GAN | 2Dt cardiac MRI enhancement | L1 fast-Fourier transform loss |
| Seq2Seq (2024) | GAN | 3D/4D MRI synthesis | L1 loss, perceptual loss, adversarial loss, cycle-consistent loss |
| DPI-MoCo (2024) | GAN | 4D CBCT reconstruction | MSE loss, GAN loss, NCC loss, smooth loss |
| DDM (2022) | Diffusion model | 4D dynamic MRI synthesis | Diffusion denoising loss, NCC loss, KL loss |
| Reynaud et al. (2023) | Diffusion model, Transformer | Echocardiography video synthesis | Diffusion denoising loss |
| HeartBeat (2024) | Diffusion model, VAE | Echocardiography video synthesis | Diffusion denoising loss |
| Endora (2024) | Diffusion model, Transformer | Endoscopy video synthesis | Diffusion denoising loss |
📈 Disease Progression Prediction
Show 12 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| Moya-Sáez et al. (2022) | CNN | Glioblastoma survival prediction | L1 loss |
| EfficientNet B0 (2024) | CNN | Hematoma expansion prediction | Focal loss |
| DaniNet (2019) | GAN | Mimic disease progression | Biological constraints loss, Deformation loss |
| GP-GAN (2020) | GAN | Brain tumor growth prediction | Adversarial loss, L1 loss, Dice loss |
| Song et al. (2023) | GAN | Longitudinal MRI prediction | Adversarial loss, Binary cross-entropy loss, Gradient difference loss |
| DCGAN and SRGAN (2024) | GAN | Alzheimer's disease progression | Adversarial loss, MSE loss, VGG Loss |
| TADM (2024) | Diffusion model | Brain neurodegenerative prediction | Diffusion denoising loss |
| BrLP (2024) | Diffusion model | Disease progression prediction | Diffusion denoising loss |
| DiffTumor (2024) | Diffusion model，VAE | Generalizable tumor synthesis | Diffusion denoising loss |
| PASTA (2025) | Diffusion model，VAE | Tumor synthesis Foundation model | Diffusion denoising loss |
| SADM (2023) | Diffusion model, AR | Longitudinal MRI Generation | Diffusion denoising loss |
| TaDiff (2025) | Diffusion model | Longitudinal MRI Generation and Glioma Growth Prediction | Diffusion denoising loss |
🏗️ Foundation Models
Show 10 entries
| Method / Publication | Model | Application | Loss |
|---|---|---|---|
| MedDiff-FM (2024) | Diffusion model | CT Foundation model | Denoising diffusion loss |
| RETFound-DE (2025) | Diffusion model | Retinal Foundation model | Denoising diffusion loss |
| RoentGen (2024) | Diffusion model | Chest X-ray Foundation model | Denoising diffusion loss |
| MINIM (2024) | Diffusion model | OCT/CT/X-ray/MRI Foundation model | Denoising diffusion loss |
| BME-X (2024) | CNN | MRI Foundation model | Cross-entropy loss，MSE loss |
| Triad (2025) | Transformer, VAE | MRI Foundation Model | L1 loss, Log-ratio loss |
| BEPH (2025) | Transformer | Pathology Foundation model | MSE loss |
| Prov-GigaPath (2024) | Transformer | Pathology Foundation model | Contrastive loss，MSE loss |
| MONET (2024) | Transformer | Image-text Foundation model | Contrastive loss，Cross-entropy loss |
| MaCo (2024) | Transformer | Radiography-reports Foundation model | InfoNCE loss，MAE loss |
🤝 Contributing
Pull requests are welcome for:
- Fixing broken links
- Adding official code or project pages
- Correcting dataset metadata or scale descriptions
- Extending the list with new generative medical imaging resources
🙏 Acknowledgments
This repository is built from the supplementary materials of our comprehensive survey. We extend our gratitude to:
- 🎓 Researchers who made their datasets and code publicly available
- 💻 Open-source community for advancing medical AI
- 🏥 Medical institutions for supporting data sharing
- 📚 Reviewers and editors for their valuable feedback
Source and Scope:
- Primary source: supplementary tables from the review paper
- Dataset entries mainly from Table S9
- Method entries mainly from Tables S2-S8 and Table S10
- Links mapped to embedded hyperlink targets in the supplementary file
- This repository is a curated index, not a claim that every linked resource is official code from the original paper authors
📜 License
- Content : CC BY 4.0 - Free to share and adapt with attribution
- Code: MIT License - Free to use, modify, and distribute
- 转载请注明出处
📖 Citation
If you find this repository helpful for your research, please consider citing our survey:
bibtex
@article{generative_ai_medical_imaging_2025,
title = {Generative Artificial Intelligence in Medical Imaging: Foundations, Progress, and Clinical Translation},
journal = {Research},
year = {2025},
doi = {10.34133/research.1029},
url = {https://spj.science.org/doi/full/10.34133/research.1029}
}