Ultrasound Image Sequences Research Articles

Contrast-enhanced ultrasound image sequences based on radiomics analysis for diagnosis of metastatic cervical lymph nodes from thyroid cancer.

Thyroid cancer (TC) prone to cervical lymph node (CLN) metastasis both before and after surgery. Ultrasonography (US) is the first-line imaging method for evaluating the thyroid gland and CLNs. However, this assessment relies mainly on the subjective judgment of the sonographer and is very much dependent on the sonographer's experience. This prospective study was designed to construct a machine learning model based on contrast-enhanced ultrasound (CEUS) videos of CLNs to predict the risk of CLN metastasis in patients with TC. Patients who were proposed for surgical treatment due to TC from August 2019 to May 2020 were prospectively included. All patients underwent US of CLNs suspected of metastasis, and a 2-minute imaging video was recorded. After target tracking, feature extraction, and feature selection through the lymph node imaging video, three machine learning models, namely, support vector machine, linear discriminant analysis (LDA), and decision tree (DT), were constructed, and the sensitivity, specificity, and accuracy of each model for diagnosing lymph nodes were calculated by leave-one-out cross-validation (LOOCV). A total of 75 lymph nodes were included in the study, with 42 benign cases and 33 malignant cases. Among the machine learning models constructed, the support vector machine had the best diagnostic efficacy, with a sensitivity of 93.0%, a specificity of 93.8%, and an accuracy of 93.3%. The machine learning model based on US video is helpful for the diagnosis of whether metastasis occurs in the CLNs of TC patients.

Multimodal risk classifier for cardiorespiratory diseases taking into account concomitant diseases and synergy effect

The purpose of the research is to develop a methodology for classifying complexly structured halftone images based on a multimodal approach using methods of morphological analysis, spectral analysis and neural network modeling.Methods. A method for classifying the contours of the boundaries of segments of a complexly structured image is described. The method is based on the fact that in chronic diseases of the pancreas, there is a violation of the integrity of the contour of its border and its waviness increases due to retractions and bulges caused by an alterative inflammatory process. The method includes the stages of normalization of ultrasound images and image segmentation with the selection of the contour of the object of interest. To classify the contour of a segment boundary, it is proposed to use Fourier analysis and neural network technologies. The method is illustrated using the example of classifying the contour of the border of the pancreas on its transcutaneous acoustic image.Results. Experimental studies of the proposed methods and means for classifying medical risk were carried out on diagnostic tasks according to the following classes: "chronic pancreatitis" – "without pathology". For experimental studies, video sequences of ultrasound images of the pancreas provided by an endoscopist were used. The purpose of the experimental studies was to analyze the classification quality indicators of image classifiers with class segments "Chronic pancreatitis" and "Without pathology". The training sample of video images (frames of video sequences) included 200 examples, one hundred from each class. The quality indicator "Sensitivity" of classification for two classes is 85,7%, the indicator "Specificity" is 87,1%.Conclusion. The use of the contour analysis method in classifiers of ultrasound images of the pancreas opens up new opportunities for accessible and objective diagnosis of pancreatic diseases, expanding the capabilities of intelligent clinical decision support systems.

Volumetric Ultrasound Imaging for the Whole Soft Tissue: Toward Enhanced Thyroid Disease Examination

Ultrasound imaging (USI) is the gold standard in the clinical diagnosis of thyroid diseases. Compared with two-dimensional (2D) USI, three-dimensional (3D) USI could provide more structural information. However, the unstable pressure generated by the hand-hold ultrasound probe scanning can cause tissue deformation, especially in soft tissues such as the thyroid. The deformation is manifested as tissue structure being compressed in 2D USI, which results in structural discontinuity in 3D USI. Furthermore, multiple scans apply pressure in different directions to the tissue, which will cause relative displacement between the 3D images obtained from multiple thyroid scans. In this work, we proposed a framework to minimize the influence of the variation of pressure in thyroid 3D USI. To correct pressure artifacts in a single scanning sequence, an adaptive method to smooth the position of the 2D ultrasound (US) image sequence is adopted before performing volumetric reconstruction. To build a whole 3D US image including both sides of the thyroid gland, an iterative closest point (ICP) based registration pipeline is adopted to eliminate the relative displacement caused by different pressure directions. Our proposed method was validated by in vivo experiments, including healthy volunteers and volunteers with thyroid nodules at different grading levels. The thyroid gland and nodule are rendered intelligently in the whole scanning region to facilitate the observation of 3D USI results by the doctor. This work might make a positive contribution to the clinical diagnosis of diseases of the thyroid or other soft tissues.

Super-Resolution Ultrasound Imaging for Monitoring the Therapeutic Efficacy of a Vascular Disrupting Agent in an Animal Model of Breast Cancer.

Evaluate the use of super-resolution ultrasound (SRUS) imaging for the early detection of tumor response to treatment using a vascular-disrupting agent (VDA). A population of 28 female nude athymic mice (Charles River Laboratories) were implanted with human breast cancer cells (MDA-MB-231, ATCC) in the mammary fat pad and allowed to grow. Ultrasound imaging was performed using a Vevo 3100 scanner (FUJIFILM VisualSonics Inc) equipped with the MX250 linear array transducer immediately before and after receiving bolus injections of a microbubble (MB) contrast agent (Definity, Lantheus Medical Imaging) via the tail vein. Following baseline ultrasound imaging, VDA drug (combretastatin A4 phosphate, CA4P, Sigma Aldrich) or control saline was injected via the placed catheter. After 4 or 24 hours, repeat ultrasound imaging along the same tumor cross-section occurred. Direct intratumoral pressure measurements were obtained using a calibrated sensor. All raw ultrasound data were saved for offline processing and SRUS image reconstruction using custom MATLAB software (MathWorks Inc). From a region encompassing the tumor space and the entire postprocessed ultrasound image sequence, time MB count (TMC) curves were generated in addition to traditional SRUS maps reflecting MB enumeration at each pixel location. Peak enhancement (PE) and wash-in rate (WIR) were extracted from these TMC curves. At termination, intratumoral microvessel density (MVD) was quantified using tomato lectin labeling of patent blood vessels. SRUS images exhibited a clear difference between control and treated tumors. While there was no difference in any group parameters at baseline (0 hour, P > .09), both SRUS-derived PE and WIR measurements in tumors treated with VDA exhibited significant decreases by 4 (P = .03 and P = .05, respectively) and 24 hours (P = .02 and P = .01, respectively), but not in control group tumors (P > .22). Similarly, SRUS derived microvascular maps were not different at baseline (P = .81), but measures of vessel density were lower in treated tumors at both 4 and 24 hours (P < .04). An inverse relationship between intratumoral pressure and both PE and WIR parameters were found in control tumors (R2 > .09, P < .03). SRUS imaging is a new modality for assessing tumor response to treatment using a VDA.

Abstract PO2-28-10: Analysis of whole breast ultrasound image sequence reading time with a new FDA approved automated whole breast ultrasound tomography system for supplemental screening in women with dense breasts

Abstract Purpose: Ultrasound technology as a supplement to mammography for screening women with dense breasts increases cancer detection. However, automated whole breast ultrasound technologies have been shown to substantially increase radiologist reading time. We conducted a reading validation program to assess how quickly and accurately radiologists could interpret images from a new automated whole breast ultrasound prior to launching a prospective case collection registry. Methods: The study analyzed 500 reading times of bilateral screening breast ultrasounds obtained from SoftVue Automated Whole Breast Ultrasound (SV) (Delphinus Medical) in women with dense breasts. For this proctored reading validation program, 20 bilateral screening ultrasound cases were selected. Thecases included a mix of normal, cysts, benign masses, cancers, and implants. 25 MSQA-trained radiologists with varying experience from 9 institutions across the United States each read the 20 cases. The same hanging protocol was used, beginning with sequences of Wafer and Sound Speed to identify an area of interest, Reflection to determine if the area persists, and then Stiffness Fusion for further confirmation and characterization. The primary outcome was reading time per bilateral ultrasound. Reading times were compared using linear mixed effects models for findings in neither, one, or both breasts and for subgroups of cancer and benign findings. Mean reading time was 3.3± 1.76 minutes (n=498, median 2.9min) per case. Cases that were negative in both breasts tended to have faster read times (2.7min, sd=1.32) compared to those with findings in one breast (3.2min, sd=1.57) or findings in both breasts (3.8min, sd=2.14). Mean read times for cases with clinical findings were within a minute of the mean time for normal cases across all case types. Conclusion Breast image sequences of Wafer, Sound Speed, Reflection, and Stiffness Fusion from SV ultrasound tomography facilitate bilateral ultrasound mean reading times of less than 4 minutes per case. Clinical Significance: SV, used as an adjunct to mammography in screening women with dense breasts, offers the benefits of increased cancer detection with supplemental ultrasound with little increase in radiologist reading time. Table 1: Analysis of Bilateral SV Read Time Across Types of Case Clinical Findings Citation Format: Mary Yamashita, Taylor Mahoney, Patrick Walker, Linda Hovanessian Larsen. Analysis of whole breast ultrasound image sequence reading time with a new FDA approved automated whole breast ultrasound tomography system for supplemental screening in women with dense breasts [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO2-28-10.

Automated Classification of Abdominal Ultrasound lmages of the Pancreas Based on the Spectral Representation of the Border’s Contours

The purpose of the research is to develop a methodology for classifying complexly structured halftone images based on a multimodal approach using methods of morphological analysis, spectral analysis and neural network modeling.Methods. А method for classifying the contours of the boundaries of segments of a complexly structured image is described. Тhe method is based on the fact that in chronic diseases of the pancreas, there is a violation of the integrity of the contour of its border and its waviness increases due to retractions and bulges caused by an alterative inflammatory process. Тhe method includes the stages of normalization of ultrasound images and image segmentation with the selection of the contour of the object of interest. Тo classify the contour of a segment boundary, it is proposed to use Fourier analysis and neural network technologies. Тhe method is illustrated using the example of classifying the contour of the border of the pancreas on its transcutaneous acoustic image.Results. Еxperimental studies of the proposed methods and means for classifying medical risk were carried out on diagnostic tasks according to the following classes: "chronic pancreatitis" – "without pathology". For experimental studies, video sequences of ultrasound images of the pancreas provided by an endoscopist were used. Тhe purpose of the experimental studies was to analyze the classification quality indicators of image classifiers with class segments "Chronic pancreatitis" and "Without pathology". Тhe training sample of video images (frames of video sequences) included 200 examples, one hundred from each class. Тhe quality indicator "Sensitivity" of classification for two classes is 85,7%, the indicator "Specificity" is 87,1%.Тhe use of the contour analysis method in classifiers of ultrasound images of the pancreas opens up new opportunities for accessible and objective diagnosis of pancreatic diseases, expanding the capabilities of intelligent clinical decision support systems.

Optimization of classifying accurate and misarticulated speech sounds for use in a gamified real-time ultrasound biofeedback system

Integrating ultrasound biofeedback therapy (UBT) into a real-time, gamified interface to provide articulatory feedback for speech remediation promotes an external focus of attention, thereby reducing the complex cognitive demands required for standard UBT. Previous studies have shown that accuracy of American English rhotic /r/ can be predicted by a single parameter, the difference between tongue dorsum and blade displacements measured by ultrasound imaging during speech production. This parameter has classified speech productions of rhotic syllables as correct versus misarticulated with a classification accuracy up to 85%. However, implementation of this classification approach into real-time gamified UBT, including both measurement timing and establishment of difficulty levels for progressive therapy, would benefit from optimization using a larger dataset. 2,450 productions of 10 distinct rhotic syllables (including prevocalic and postvocalic contexts) from 50 children, with and without misarticulations, were analyzed. For each production, ultrasound image sequences were processed by TonguePART software to acquire tongue displacement trajectories, and accuracy was judged by trained listeners using a visual analog scale. Analyses were conducted to optimize selection of the image frame for classification, determine parameter thresholds appropriate for real-time prediction of /r/ production accuracy, and integrate these thresholds into a difficulty level design for gamified UBT.

Volumetric B-mode ultrasound and Doppler Imaging: Automatic Tracking With One Single Camera.

Vascular diseases may occur in the upper extremities, and the lesions can span the entire length of the blood vessel. One of the most popular methods to identify vascular disorders is ultrasound Doppler imaging. However, traditional two-dimensional (2D) ultrasound Doppler imaging cannot capture the entire length of a long vessel in one image. Medical professionals often have to painstakingly reconstruct three-dimensional (3D) data using 2D ultrasound images to locate the lesions, especially for large blood vessels. 3D ultrasound Doppler imaging can display the morphological structure of blood vessels and the distribution of lesions more directly, providing a more comprehensive view compared to 2D imaging. In this work, we propose a wide-range 3D volumetric ultrasound Doppler imaging system with dual modality, in which a high-definition camera is adopted to automatically track the movement of the ultrasound transducer, simultaneously capturing a corresponding sequence of 2D ultrasound Doppler images. We conducted experiments on human arms using our proposed system and separately with X-ray computerized tomography (X-CT). The comparison results prove the potential value of our proposed system in the diagnosis of arm vascular diseases.

Standardized acquisition and documentation of cine loops on conventional thyroid ultrasound

Ultrasound is the basic imaging method for the assessment of the thyroid gland. Due to the high prevalence of structural disease, the examination procedure is used very frequently in Germany, in many cases in the context of follow-up. The assessment of thyroid pathologies and their dynamics is subjected to relevant inter- and intraobserver variability. Findings that were not identified during live ultrasound cannot be assessed retrospectively. Applying an SOP for the acquisition and documentation of standardized video sequences of ultrasound images (so-called cine loops), allows for a secondary retrospective evaluation of the thyroid gland, taking into account previously acquired images analogous to other cross-sectional imaging methods such as CT or MRI. The cine loops can be acquired by non-physician personnel, stored to the local PACS and used for educational and research purposes.

Ultrasound Probe Calibration Method of Single-Wire Phantom Using Levenberg-Marquardt Algorithm

A freehand three-dimensional (3D) ultrasound system is a method of acquiring images using a 3D ultrasound probe or conventional two-dimensional (2D) ultrasound probe to give a 3D visualization of an object inside the body. Ultrasounds are used extensively in clinical applications since they are advantageous in that they do not bring dangerous radiation effects and have a low cost. However, a probe calibration method is needed to transform the coordinate position into a 3D visualization display, especially for image-guided intervention. The current ultrasound probe calibration system usually uses the numerical regression method for the N-wire phantom, which has problems in accuracy and reliability due to nonlinear point scattered ultrasound image data. Hence, a method for ultrasound probe positional calibration of single-wire phantom using the Levenberg-Marquardt algorithm (LMA) was proposed to overcome this weakness. This experiment consisted of an optical tracking system setup, a 2D ultrasound probe with marker, an ultrasound machine, and a single-wire object in a water container equipped with a marker. The position and orientation of the marker in a 2D ultrasound probe and the marker in the water container were tracked using the optical tracking system. A 2D ultrasound probe was equipped with a marker connected wirelessly using an optical tracking system to capture the single-wire object. The resulting sequences of 2D ultrasound images were reconstructed and visualized into 3D ultrasound images using three transformations, ultrasound beam to ultrasound probe’s marker, single-wire phantom position to container’s marker, and the 3D visualization transformation. The LMA was used to determine the best optimization parameters for determining the exact position and representing that 3D visualization. The experiment result showed that the lowest mean square error (MSE), rotation error, and translation error were 0.45 mm, 0.25°, and 0.3828 mm, respectively.

SequenceMorph: A Unified Unsupervised Learning Framework for Motion Tracking on Cardiac Image Sequences.

Modern medical imaging techniques, such as ultrasound (US) and cardiac magnetic resonance (MR) imaging, have enabled the evaluation of myocardial deformation directly from an image sequence. While many traditional cardiac motion tracking methods have been developed for the automated estimation of the myocardial wall deformation, they are not widely used in clinical diagnosis, due to their lack of accuracy and efficiency. In this paper, we propose a novel deep learning-based fully unsupervised method, SequenceMorph, for in vivo motion tracking in cardiac image sequences. In our method, we introduce the concept of motion decomposition and recomposition. We first estimate the inter-frame (INF) motion field between any two consecutive frames, by a bi-directional generative diffeomorphic registration neural network. Using this result, we then estimate the Lagrangian motion field between the reference frame and any other frame, through a differentiable composition layer. Our framework can be extended to incorporate another registration network, to further reduce the accumulated errors introduced in the INF motion tracking step, and to refine the Lagrangian motion estimation. By utilizing temporal information to perform reasonable estimations of spatio-temporal motion fields, this novel method provides a useful solution for image sequence motion tracking. Our method has been applied to US (echocardiographic) and cardiac MR (untagged and tagged cine) image sequences; the results show that SequenceMorph is significantly superior to conventional motion tracking methods, in terms of the cardiac motion tracking accuracy and inference efficiency.

Free scan real time 3D ultrasound imaging with shading artefacts removal

Ultrasound imaging (USI) is a widely adopted imaging method in clinical diagnosis owing to its low cost, convenience, and safety. However, due to the complex acoustic attenuation, two-dimensional (2D) USI lacks the capability to achieve a clear imaging result when the target is shaded by high echo tissues. This paper proposes a three-dimensional (3D) free-scan real-time ultrasound imaging (FRUSI) method. By integrating 2D ultrasound image sequences around the region of interest (ROI) with a real-time and spatially accurate probe tracking method, the proposed FRUSI system provides clear and accurate ultrasound images for medical study. The experiment results on reconstruction precision and accuracy show the potential ability of our proposed system to provide high-quality 3D ultrasound imaging. Moreover, previously shaded targets can be discerned clearly in the same scan plane in both phantom studies and in vivo studies on the human finger joint. The performance of the proposed FRUSI system has demonstrated its potential value for clinical diagnosis to provide high ultrasound imaging quality and rich details in spatial information. Due to the convenient setup, the FRUSI system might potentially be expanded to other ultrasound imaging modalities.

High frequency ultrasound vibrational shear wave elastography for preclinical research

Preclinical evaluation of novel therapies using models of cancer is an important tool in cancer research, where imaging can provide non-invasive tools to characterise the internal structure and function of tumours. The short propagation paths when imaging tumours and organs in small animals allow the use of high frequencies for both ultrasound and shear waves, providing the opportunity for high-resolution shear wave elastography and hence its use for studying the heterogeneity of tissue elasticity, where heterogeneity may be a predictor of tissue response. Here we demonstrate vibrational shear wave elastography (VSWE) using a mechanical actuator to produce high frequency (up to 1000 Hz) shear waves in preclinical tumours, an alternative to the majority of preclinical ultrasound SWE studies where an acoustic radiation force impulse is required to create a relatively low-frequency broad-band shear-wave pulse. We implement VSWE with a high frequency (17.8 MHz) probe running a focused line-by-line ultrasound imaging sequence which as expected was found to offer improved detection of 1000 Hz shear waves over an ultrafast planar wave imaging sequence in a homogenous tissue-mimicking phantom. We test the VSWE in an ex vivo tumour xenograft, demonstrating the ability to detect shear waves up to 10 mm from the contactor position at 1000 Hz. By reducing the kernel size used for shear wave speed estimation to 1 mm we are able to produce shear wave speed images with spatial resolution of this order. Finally, we present VSWE data from xenograft tumours in vivo, demonstrating the feasibility of the technique in mice under isoflurane sedation. Mean shear wave speeds in the tumours are in good agreements with those reported by previous authors. Characterising the frequency dependence of shear wave speed demonstrates the potential to quantify the viscoelastic properties of tumours in vivo.

Interframe Echo Intensity Variation of Subregions and Whole Plaque in Two-Dimensional Carotid Ultrasonography: Simulations and In Vivo Observations.

The risk of cardiovascular disease is associated with the echo intensity of carotid plaques in ultrasound images and their cardiac cycle-induced intensity variations. In this study, we aimed to 1) explore the underlying origin of echo intensity variations by using simulations and 2) evaluate the association between the two-dimensional (2D) spatial distribution of these echo intensity variations and plaque vulnerability. First, we analyzed how out-of-plane motion and compression of simulated scattering spheres of different sizes affect the ultrasound echo intensity. Next, we propose a method to analyze the features of the 2D spatial distribution of interframe plaque echo intensity in carotid ultrasound image sequences and explore their associations with plaque vulnerability in experimental data. The simulations showed that the magnitude of echo intensity changes was similar for both the out-of-plane motion and compression, but for scattering objects smaller than 1 mm radius, the out-of-plane motion dominated. In experimental data, maps of the 2D spatial distribution of the echo intensity variations had a low correlation with standard B-mode echo intensity distribution, indicating complementary information on plaque tissue composition. In addition, we found the existence of ∼1 mm diameter subregions with pronounced echo intensity variations associated with plaque vulnerability. The results indicate that out-of-plane motion contributes to intra-plaque regions of high echo intensity variation. The 2D echo intensity variation maps may provide complementary information for assessing plaque composition and vulnerability. Further studies are needed to verify this method's role in identifying vulnerable plaques and predicting cardiovascular disease risk.

Temporal enhanced Ultrasound: A new method for detection of porosity defects in composites

Non-Destructive Evaluation (NDE) methods are commonly employed for identifying porosity, which is one of the most common manufacturing defects observed in composite structures. Among current widely used approaches are conventional ultrasonic methods such as pulse-echo analysis based on loss of signal amplitude. Application of these conventional ultrasonic methods, however, can be challenging in cases where the loss of signal is negligible, such as with porosity. In this paper, we propose Temporal-enhanced Ultrasound (TeUS) as a novel ultrasound-based imaging technique for NDE of composites. TeUS represents the analysis of a sequence of ultrasound images obtained from composites by varying an image acquisition parameter, such as the focal point, over the sequence. We present details on the analytical formulation of TeUS, followed by extensive simulation and experimental results to demonstrate the efficacy of TeUS for porosity detection, compared to a conventional ultrasound-based porosity detection method. The results indicate that TeUS has excellent sensitivity for identifying subtle changes in the level of porosity within composite samples, with the possibility to reduce the need for reference samples.

Super-Resolution Ultrasound Localization Microscopy of Microvascular Structure and Flow for Distinguishing Metastatic Lymph Nodes - An Initial Human Study.

Detecting and distinguishing metastatic lymph nodes (LNs) from those with benign lymphadenopathy are crucial for cancer diagnosis and prognosis but remain a clinical challenge. A recent advance in super-resolution ultrasound (SRUS) through localizing individual microbubbles has broken the diffraction limit and tracking enabled in vivo noninvasive imaging of vascular morphology and flow dynamics at a microscopic level. In this study we hypothesize that SRUS enables quantitative markers to distinguish metastatic LNs from benign ones in patients with lymphadenopathy. Clinical contrast-enhanced ultrasound image sequences of LNs from 6 patients with lymph node metastasis and 4 with benign lymphadenopathy were acquired and motion-corrected. These were then used to generate super-resolution microvascular images and super-resolved velocity maps. From these SRUS images, morphological and functional measures were obtained including micro-vessel density, fractal dimension, mean flow speed, and Local Flow Direction Irregularity (LFDI) measuring the variance in local flow direction. These measures were compared between pathologically proven reactive and metastasis LNs. Our initial results indicate that the difference in the indicator of flow irregularity (LFDI) derived from the SRUS images is statistically significant between the two groups. The LFDI is 60% higher in metastatic LNs compared with reactive nodes. This pilot study demonstrates the feasibility of super-resolution ultrasound for clinical imaging of lymph nodes and the potential of using the irregularity of local blood flow directions afforded by SRUS for the characterization of LNs.

Extending the pixel difference metric from 2D to 3D/4D ultrasound

It is frequently of interest to examine differences (e.g., the 1st or 2nd difference) in signals and images. Differencing can be employed for the detection of edges in images, or the detection of events in temporal sequences. One such event is the onset of articulatory movement in ultrasound image sequences of the tongue. It has been quantified using a form of Euclidean distance between successive frames of image scan line data, referred to as Pixel Difference. When applied to 3-D/4-D ultrasound data of the tongue (after scan conversion), Pixel Difference exhibited a surprisingly poor signal-to-noise ratio. In this presentation, we test the hypotheses that (1) scan conversion amplifies the effect of speckle noise near the top of the ultrasound image (i.e. above the tongue), and (2) the Pixel Difference signal-to-noise ratio in 3-D/4-D ultrasound differs from 2-D ultrasound by a power of 3/2 due to the added spatial dimension.

Multi3: multi-templates siamese network with multi-peaks detection and multi-features refinement for target tracking in ultrasound image sequences

Objective. Radiation therapy requires a precise target location. However, respiratory motion increases the uncertainties of the target location. Accurate and robust tracking is significant for improving operation accuracy. Approach. In this work, we propose a tracking framework Multi3, including a multi-templates Siamese network, multi-peaks detection and multi-features refinement, for target tracking in ultrasound sequences. Specifically, we use two templates to provide the location and deformation of the target for robust tracking. Multi-peaks detection is applied to extend the set of potential target locations, and multi-features refinement is designed to select an appropriate location as the tracking result by quality assessment. Main results. The proposed Multi3 is evaluated on a public dataset, i.e. MICCAI 2015 challenge on liver ultrasound tracking (CLUST), and our clinical dataset provided by the Chinese People’s Liberation Army General Hospital. Experimental results show that Multi3 achieves accurate and robust tracking in ultrasound sequences (0.75 ± 0.62 mm and 0.51 ± 0.32 mm tracking errors in two datasets). Significance. The proposed Multi3 is the most robust method on the CLUST 2D benchmark set, exhibiting potential in clinical practice.

Segmentation of ultrasound image sequences by combing a novel deep siamese network with a deformable contour model

Segmentation of ultrasound image sequences by combing a novel deep siamese network with a deformable contour model

Automatic Tracking of Muscle Fiber Direction in Ultrasound Images Based on Improved Kalman Filter

Ultrasound myograph (SMG) is a real-time and dynamic acquisition of muscle structure parameter changes by recording ultrasound images of muscle contraction through an ultrasound instrument. Muscle parameters are essential for judging whether the muscle and the human body are healthy. In order to solve the problem of muscle fiber tracking in a sequence of ultrasound muscle images, we propose a method to track the direction of muscle fibers automatically based on the improved Kalman filter. Firstly, the measurement value of the muscle fiber direction is obtained by introducing a reference line into the ultrasound muscle image based on deep learning. Secondly, the framework of a Kalman filter is improved by introducing a set of neural units. Finally, the optimal estimated value of muscle fiber direction is obtained by combining the measured value with the improved Kalman filter. It is verified by conducting experiments where the result obtained by our proposed method is closer to the manually labeled value compared with the original measurement method, and the root mean square error is reduced by about 10%.