⚠️ 以下所有内容总结都来自于 大语言模型的能力，如有错误，仅供参考，谨慎使用
🔴 请注意：千万不要用于严肃的学术场景，只能用于论文阅读前的初筛！
💗 如果您觉得我们的项目对您有帮助 ChatPaperFree ，还请您给我们一些鼓励！⭐️ HuggingFace免费体验
2025-08-14 更新
PADReg: Physics-Aware Deformable Registration Guided by Contact Force for Ultrasound Sequences
Authors:Yimeng Geng, Mingyang Zhao, Fan Xu, Guanglin Cao, Gaofeng Meng, Hongbin Liu
Ultrasound deformable registration estimates spatial transformations between pairs of deformed ultrasound images, which is crucial for capturing biomechanical properties and enhancing diagnostic accuracy in diseases such as thyroid nodules and breast cancer. However, ultrasound deformable registration remains highly challenging, especially under large deformation. The inherently low contrast, heavy noise and ambiguous tissue boundaries in ultrasound images severely hinder reliable feature extraction and correspondence matching. Existing methods often suffer from poor anatomical alignment and lack physical interpretability. To address the problem, we propose PADReg, a physics-aware deformable registration framework guided by contact force. PADReg leverages synchronized contact force measured by robotic ultrasound systems as a physical prior to constrain the registration. Specifically, instead of directly predicting deformation fields, we first construct a pixel-wise stiffness map utilizing the multi-modal information from contact force and ultrasound images. The stiffness map is then combined with force data to estimate a dense deformation field, through a lightweight physics-aware module inspired by Hooke’s law. This design enables PADReg to achieve physically plausible registration with better anatomical alignment than previous methods relying solely on image similarity. Experiments on in-vivo datasets demonstrate that it attains a HD95 of 12.90, which is 21.34% better than state-of-the-art methods. The source code is available at https://github.com/evelynskip/PADReg.
超声形变配准估计两对形变超声图像之间的空间变换，这对于捕捉甲状腺结节和乳腺癌等疾病的生物力学特性并提高诊断准确性至关重要。然而，超声形变配准仍然是一项巨大挑战，尤其是在大变形情况下。超声图像中固有的低对比度、强噪声和模糊的组织边界严重阻碍了可靠的特征提取和对应关系匹配。现有方法常常存在解剖结构对齐不佳和缺乏物理可解释性的问题。为了解决这一问题，我们提出了PADReg，这是一个受接触力引导的物理感知形变配准框架。PADReg利用机器人超声系统测量的同步接触力作为物理先验来约束配准。具体而言，我们不是直接预测变形场，而是首先利用接触力和超声图像的多模态信息构建像素级刚度图。然后，将刚度图与力数据相结合，通过受胡克定律启发的轻量化物理感知模块来估计密集的变形场。这种设计使PADReg能够实现物理上合理的配准，与仅依赖图像相似性的以前的方法相比，获得了更好的解剖结构对齐。在真实数据集上的实验表明，其HD95达到12.90，比最先进的方法提高了21.34%。源代码可在https://github.com/evelynskip/PADReg找到。
论文及项目相关链接
PDF This work has been submitted to the IEEE for possible publication
Summary：
该文本介绍了超声形变注册在捕捉生物力学特性、提高甲状腺结节和乳腺癌等疾病的诊断准确性方面的重要性。针对超声形变注册在大变形下的挑战，提出了一种结合接触力测量的物理感知形变注册框架PADReg。PADReg通过构建像素级刚度图并结合力数据估计密集形变场，实现物理合理的注册，比仅依赖图像相似性的方法更好地实现解剖对齐。
Key Takeaways：
- 超声形变注册对于捕捉生物力学特性、提高疾病诊断准确性至关重要。
- 现有超声形变注册方法面临大变形下的挑战，如低对比度、噪声和模糊的组织边界。
- PADReg是一种物理感知的形变注册框架，利用接触力作为物理先验进行约束。
- PADReg通过构建像素级刚度图并结合力数据估计密集形变场，实现物理合理的注册。
- PADReg在活体数据集上的实验结果显示，其性能优于当前主流方法，高清度达到12.90，提升了21.34%。
- PADReg源码可通过公开链接获取。
点此查看论文截图
Stochastic Reconstruction of the Speed of Sound in Breast Ultrasound Computed Tomography with Phase Encoding in the Frequency Domain
Authors:Luca A. Forte
The framework of ultrasound computed tomography (USCT) has recently re-emerged as a powerful, safe and operator-independent way to image the breast. State of the art image reconstruction methods are performed with iterative techniques based on deterministic optimization algorithms in the frequency domain in the 300 kHz - 1 MHz bandwidth. Alternative algorithms with deterministic and stochastic optimization have been considered in the time-domain. In this paper, we present the equivalent stochastic inversion in the frequency domain (phase encoding), with a focus on reconstructing the speed of sound. We test the inversion algorithm on synthetic data in 2D and 3D, by explicitly differentiating between inverse crime and non-inverse crime scenarios, and compare against the deterministic inversion. We then show the results of the stochastic inversion in the frequency domain on experimental data. By leveraging on the concepts of multiple super-shots and stochastic ensembles, we provide robust evidence that image quality of a stochastic reconstruction of the speed of sound with phase encoding in the frequency domain is comparable, and essentially equivalent, to the one of a deterministic reconstruction, with the further benefit of drastically reducing reconstruction times by more than half.
超声计算机层析成像（USCT）框架最近重新出现为一种强大、安全且操作者独立的乳腺成像方法。最新的图像重建方法采用基于频率域的确定性优化算法的迭代技术，在300 kHz至1 MHz带宽范围内进行。时间域中考虑了具有确定性和随机性优化算法。在本文中，我们介绍了频率域中的等效随机反演（相位编码），重点研究声速重建。我们在二维和三维合成数据上测试了反演算法，通过明确区分逆犯罪和非逆犯罪场景，并与确定性反演进行了比较。然后，我们展示了在试验数据上应用频率域随机反演的结果。通过利用多重超炮和随机集合的概念，我们提供了强有力的证据表明，在频率域中使用相位编码进行声速的随机重建的图像质量可与确定性重建相当，且基本上等同于确定性重建的图像质量，同时能大幅减少重建时间一半以上。
论文及项目相关链接
Summary
超声计算机断层扫描（USCT）技术重新兴起，成为乳腺成像的一种强大、安全且操作者独立的方式。当前先进的图像重建方法采用频率域的确定性优化算法，在300 kHz至1 MHz的带宽范围内进行迭代技术。本文介绍了一种频率域的等效随机反演（相位编码），重点研究声波速度的重建。通过对合成数据进行测试，并与确定性反演进行比较，结果显示频率域的随机反演在图像质量方面与确定性反演相当，并进一步减少了重建时间超过一半。
Key Takeaways
- 超声计算机断层扫描（USCT）是乳腺成像的一种强大、安全、操作者独立的方式。
- 当前图像重建方法采用频率域的确定性优化算法。
- 介绍了频率域的等效随机反演（相位编码）方法，用于重建声波速度。
- 对合成数据进行的测试显示随机反演与确定性反演的图像质量相当。
- 随机反演进一步减少了重建时间。
- 通过多重超级射击和随机集合的概念，为实验结果提供了稳健的证据。
点此查看论文截图
