mlsanityData to 14 Oct 2026

Ultrasound Image-to-Video Synthesis via Latent Dynamic Diffusion Models

arXiv 2503.14966 ↗First posted 19 Mar 2025cs.CV eess.IVcited by 12

On the map: Diffusion models → Medical imaging and segmentation

Abstract (arXiv)

Ultrasound video classification enables automated diagnosis and has emerged as an important research area. However, publicly available ultrasound video datasets remain scarce, hindering progress in developing effective video classification models. We propose addressing this shortage by synthesizing plausible ultrasound videos from readily available, abundant ultrasound images. To this end, we introduce a latent dynamic diffusion model (LDDM) to efficiently translate static images to dynamic sequences with realistic video characteristics. We demonstrate strong quantitative results and visually appealing synthesized videos on the BUSV benchmark. Notably, training video classification models on combinations of real and LDDM-synthesized videos substantially improves performance over using real data alone, indicating our method successfully emulates dynamics critical for discrimination. Our image-to-video approach provides an effective data augmentation solution to advance ultrasound video analysis. Code is available at https://github.com/MedAITech/U_I2V.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Placed in Medical imaging and segmentation, one of 14 directions in Diffusion models. Builds on 23 of its 30 references that are in the map. First posted on arXiv 19 Mar 2025.

Builds on

23 of 30 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023Echo from noise: synthetic ultrasound image generation using diffusion models for real image segmentation9 May
  2. 2023Feature-Conditioned Cascaded Video Diffusion Models for Precise Echocardiogram Synthesis22 Mar

Other references · 21

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  4. 2018Towards Accurate Generative Models of Video: A New Metric & Challengescited by 1,234
  5. 2018TSM: Temporal Shift Module for Efficient Video Understandingcited by 856
  6. 2019Deep Learning for Segmentation using an Open Large-Scale Dataset in 2D Echocardiographycited by 244
  7. 2016Generating Images with Perceptual Similarity Metrics based on Deep Networkscited by 222
  8. 2020TDN: Temporal Difference Networks for Efficient Action Recognitioncited by 183
  9. 2018Stochastic Adversarial Video Predictioncited by 173
  10. 2015Attribute2Image: Conditional Image Generation from Visual Attributescited by 93
  11. 2021Stochastic Image-to-Video Synthesis using cINNscited by 49
  12. 2022A New Dataset and A Baseline Model for Breast Lesion Detection in Ultrasound Videoscited by 32
9 more
  1. 2022D'ARTAGNAN: Counterfactual Video Generationcited by 28
  2. 2023Shifting More Attention to Breast Lesion Segmentation in Ultrasound Videoscited by 16
  3. 2017Simulating Patho-realistic Ultrasound Images using Deep Generative Networks with Adversarial Learningcited by 13
  4. 2020Synthesis and Edition of Ultrasound Images via Sketch Guided Progressive Growing GANscited by 6
  5. 2022Active Token Mixercited by 6
  6. 2022Weakly-supervised High-fidelity Ultrasound Video Synthesis with Feature Decouplingcited by 6
  7. 2021Realistic Ultrasound Image Synthesis for Improved Classification of Liver Diseasecited by 5
  8. 2023Boosting Breast Ultrasound Video Classification by the Guidance of Keyframe Feature Centerscited by 5
  9. 2023A Spatial-Temporal Deformable Attention based Framework for Breast Lesion Detection in Videoscited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 8; 2026: 4.

12 papers.

  1. 2025Taming Preference Mode Collapse via Directional Decoupling Alignment in Diffusion Reinforcement Learningcited by 18
  2. 2025FB-Diff: Fourier Basis-guided Diffusion for Temporal Interpolation of 4D Medical Imagingcited by 6
  3. 2025EchoFlow: A Foundation Model for Cardiac Ultrasound Image and Video Generationcited by 5
  4. 2025Temporal Differential Fields for 4D Motion Modeling via Image-to-Video Synthesiscited by 3
  5. 2025Phenotype-Guided Generative Model for High-Fidelity Cardiac MRI Synthesis: Advancing Pretraining and Clinical Applicationscited by 2
  6. 2025Lightweight Physics-Aware Zero-Shot Ultrasound Plane-Wave Denoisingcited by 1
  7. 2025COME: Dual Structure-Semantic Learning with Collaborative MoE for Universal Lesion Detection Across Heterogeneous Ultrasound Datasetscited by 1
  8. 2026ECGFlowCMR: Pretraining with ECG-Generated Cine CMR Helps Cardiac Disease Classification and Phenotype Predictioncited by 1
  9. 2025Label-free Motion-Conditioned Diffusion Model for Cardiac Ultrasound Synthesiscited by 0
  10. 2026Clinically-Grounded Counterfactual Reasoning for Medical Video Diagnosiscited by 0
  11. 2026Low-Rank Velocity Fields as a Structural Prior for Unsupervised 4D Medical Image Interpolationcited by 0
  12. 2026Phase-Aligned Finite-Fourier Periodic Deformation for 4D Medical Image Interpolationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Sequential Posterior Sampling with Diffusion Models
  2. 2025Ultrasound Image Generation using Latent Diffusion Models
  3. 2023Exploring the Hyperparameter Space of Image Diffusion Models for Echocardiogram Generation
  4. 2024Diffusion as Sound Propagation: Physics-inspired Model for Ultrasound Image Generation
  5. 2025Enhancing Fetal Plane Classification Accuracy with Data Augmentation Using Diffusion Models
  6. 2026Feature-Space Guided Diffusion for Realistic Ultrasound Image Synthesis
  7. 2024Explainable and Controllable Motion Curve Guided Cardiac Ultrasound Video Generation
  8. 2023Diffusion Reconstruction of Ultrasound Images with Informative Uncertainty