基于EDM2的高分辨率合成胎儿超声成像基础生成模型,源自开放数据集
A Foundational EDM2-Based Generative Model for High-Resolution Synthetic Fetal Ultrasound Imaging from Open Datasets
浏览论文内容
中文总结 AI 辅助
该研究针对胎儿超声AI数据稀缺问题,提出基于EDM2的合成框架,生成512×512图像,在分类任务中准确率达93.36%,优于真实数据训练结果,为胎儿超声AI提供开放资源。
中文摘要 AI 辅助
产前超声成像是评估胎儿健康的关键手段,但AI进展受限于稀缺、隐私受限且难以标注的数据集。我们提出一种基于EDM2扩散架构的高分辨率胎儿超声合成框架,在多个公开数据集上训练,可生成涵盖6种解剖类别的512×512图像。我们的方法实现了更优的图像质量,具有更低的FID分数,且增强了下游胎儿平面分类,在微调后达到93.36%的集成准确率,优于仅用真实数据训练的结果。由一名拥有10年以上经验的资深胎儿超声专家对100张图像进行的临床评估显示,平均真实感评分为2.67/5,真实图像评分高于合成图像,存在的伪影包括平滑、散斑不规则及解剖结构不一致。可在此https URL获取用于复现本研究的代码、数据、模型及其他资源。
英文摘要
Prenatal ultrasound imaging is key for assessing fetal health, but AI progress is limited by scarce, privacy-restricted, and hard-to-annotate datasets. We propose a high-resolution fetal ultrasound synthesis framework based on the EDM2 diffusion architecture, trained on multiple public datasets to generate 512x512 images across six anatomical classes. Our method achieved improved image quality with lower FID scores and enhanced downstream fetal plane classification, reaching 93.36% ensemble accuracy after fine-tuning, surpassing real-data-only training. Clinical evaluation by an experienced fetal ultrasound specialist (10+ years) on 100 images yielded a mean realism score of 2.67/5, with real images rated higher than synthetic. Artefacts included smoothing, speckle irregularities, and anatomical inconsistencies. Code, data, models and other resources to reproduce this work are available at https://github.com/xfetus/fetal-ultrasound-edm2.
发表机构
- University of Southampton(南安普顿大学)
- Tsinghua University(清华大学)
- King’s College London(伦敦国王学院)
- University College London(伦敦大学学院)
机构由 AI 辅助整理,请以论文原文为准。