In recent years, deep learning (DL) has been successfully applied to the analysis and processing of ultrasound images. To date, most of this research has focused on segmentation and view recognition. This article benchmarks different convolutional neural network algorithms for motion estimation in ultrasound imaging. We evaluated and compared several networks derived from FlowNet2, one of the most efficient architectures in computer vision. The networks were tested with and without transfer learning, and the best configuration was compared against the particle imaging velocimetry method, a popular state-of-the-art block-matching algorithm. Rotations are known to be difficult to track from ultrasound images due to a significant speckle decorrelation. We thus focused on the images of rotating disks, which could be tracked through speckle features only. Our database consisted of synthetic and in vitro B-mode images after log compression and covered a large range of rotational speeds. One of the FlowNet2 subnetworks, FlowNet2SD, produced competitive results with a motion field error smaller than 1 pixel on real data after transfer learning based on the simulated data. These errors remain small for a large velocity range without the need for hyperparameter tuning, which indicates the high potential and adaptability of DL solutions to motion estimation in ultrasound imaging.

中文翻译：

---

卷积神经网络用于超声图像运动估计的先导研究。

近年来，深度学习（DL）已成功应用于超声图像的分析和处理。迄今为止，大多数研究都集中在分割和视图识别上。本文对用于超声成像运动估计的不同卷积神经网络算法进行了基准测试。我们评估并比较了源自FlowNet2的几种网络，FlowNet2是计算机视觉中最高效的体系结构之一。在有无转移学习的情况下对网络进行了测试，并且将最佳配置与粒子成像测速方法（一种流行的最先进的块匹配算法）进行了比较。已知由于明显的斑点去相关而难以从超声图像跟踪旋转。因此，我们专注于旋转磁盘的图像，只能通过斑点特征进行跟踪。我们的数据库由合成和体外B模式对数压缩后的图像，并涵盖了很大的旋转速度范围。在基于模拟数据进行转移学习之后，FlowNet2子网之一FlowNet2SD产生了具有竞争力的结果，其真实数据的运动场误差小于1个像素。对于较大的速度范围，这些误差仍然很小，无需进行超参数调整，这表明DL解决方案具有很高的潜力，并且可以很好地适应超声成像中运动估计的需求。