Ultrasound Speckle Tracking with Deep Convolutional Neural Network

Ultrasonic imaging is convenient and safe for cardiovascular disease diagnosis. Speckle tracking can obtain accurate myocardial movement data and provide important information for the diagnosis of cardiac function. Block matching method and optical flow method are the most commonly used speckle tracking methods. However, the accuracy of these methods cannot meet the needs of clinical application. Deep learning is applied to speckle tracking technology. Based on the correlation filters given to the deep convolution network, the migration learning method is introduced to obtain the feature mapping on the convolution layer on the pre-trained ImageNet VGG19 network. The feature mapping is used as the training data of correlation filters, and the tracking results obtained from convolution layers with different depths are filtered frame by frame, giving different weights to obtain the optimal tracking position within a certain search range. Then the correlation filter is updated to track the myocardial motion. The proposed deep learning based method has better accuracy for myocardial motion tracking, which indicates that the target tracking method based on convolutional neural network has potential advantages in this field.

Neural Multi-Scale Self-Supervised Registration for Echocardiogram Dense Tracking

Echocardiography has become routinely used in the diagnosis of cardiomyopathy and abnormal cardiac blood flow. However, manually measuring myocardial motion and cardiac blood flow from echocar-diogram is time-consuming and error-prone. Computer algorithms that can automatically track and quantify myocardial motion and cardiac blood flow are highly sought after, but have not been very successful due to noise and high variability of echocardiography. In this work, we propose a neural multi-scale self-supervised registration (NMSR) method for automated myocardial and cardiac blood flow dense tracking. NMSR incorporates two novel components: 1) utilizing a deep neural net to parameterize the velocity field between two image frames, and 2) optimizing the parameters of the neural net in a sequential multi-scale fashion to account for large variations within the velocity field. Experiments demonstrate that NMSR yields significantly better registration accuracy than the state-of-the-art methods, such as advanced normalization tools (ANTs) and Voxel Morph, for both myocardial and cardiac blood flow dense tracking. Our approach promises to provide a fully automated method for fast and accurate analyses of echocardiograms.