针对超声图像纹理特性，提出新框架提升医学影像模型泛化能力
A texture-based framework for foundational ultrasound models
- 将自监督学习重构为纹理分析问题，捕捉超声特有物理特征
- 在多个临床任务中表现超越大模型，如脊髓血肿检测准确率100%
- 适合医疗AI研究者与超声诊断系统开发者使用
超声成像是最广泛应用的医学影像技术，但其图像由组织散射、反射及声速变化决定，形成独特且受限的纹理模式，与自然图像统计特性显著不同。这使得为自然图像设计的算法难以适用于超声。尽管领域正转向基础模型以利用其泛化能力，但现有模型因未考虑超声物理机制而性能受限。本文提出纹理超声语义分析（TUSA），将自监督学习重新定义为纹理分析问题，通过可扩展的对比学习方法从简单B-mode图像中提取真实域特定表征。我们在开源、模拟和活体数据组合上训练模型，并与多个更大基础模型比较，结果表明TUSA在困难下游任务上具备更优泛化能力，包括在线数据集和本研究采集的临床眼部数据集。模型在检测新冠（70%）、脊髓血肿（100%）和玻璃体出血（97%）方面表现优异，且与肝脂肪变性（r=0.83）、射血分数（r=0.63）和血氧饱和度（r=0.38）等定量参数高度相关。代码与权重已开源。
原文摘要 · Abstract (English)
Ultrasound is the most widely used medical imaging modality, yet the images it produces are fundamentally unique, arising from tissue-dependent scattering, reflection, and speed-of-sound variations that produce a constrained set of characteristic textures that differ markedly from natural-image statistics. These acoustically driven patterns make ultrasound challenging for algorithms originally designed for natural images. To bridge this gap, the field has increasingly turned to foundation models, hoping to leverage their generalization capabilities. However, these models often falter in ultrasound applications because they are not designed for ultrasound physics, they are merely trained on ultrasound data. Therefore, it is essential to integrate ultrasound-specific domain knowledge into established learning frameworks. We achieve this by reformulating self-supervised learning as a texture-analysis problem, introducing texture ultrasound semantic analysis (TUSA). Using TUSA, models learn to leverage highly scalable contrastive methods to extract true domain-specific representations directly from simple B-mode images. We train a TUSA model on a combination of open-source, simulated, and in vivo data. The latent space is compared to several larger foundation models, demonstrating that our approach gives TUSA models better generalizability for difficult downstream tasks on unique online datasets as well as a clinical eye dataset collected for this study. Our model achieves higher accuracy in detecting COVID (70%), spinal hematoma (100%) and vitreous hemorrhage (97%) and correlates more closely with quantitative parameters like liver steatosis (r = 0.83), ejection fraction (r = 0.63), and oxygen saturation (r = 0.38). We open-source the model weights and training script: https://github.com/talg2324/tusa
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。