Left Ventricular Segmentation, Warping, and Myocardial Registration for Automated Strain Measurement.

The left ventricular global longitudinal strain (LVGLS) is a crucial prognostic indicator. However, inconsistencies in measurements due to the speckle tracking algorithm and manual adjustments have hindered its standardization and democratization. To solve this issue, we proposed a fully automated strain measurement by artificial intelligence-assisted LV segmentation contours. The LV segmentation model was trained from echocardiograms of 368 adults (11,125 frames). We compared the registration-like effects of dynamic time warping (DTW) with speckle tracking on a synthetic echocardiographic dataset in experiment-1. In experiment-2, we enrolled 80 patients to compare the DTW method with commercially available software. In experiment-3, we combined the segmentation model and DTW method to create the artificial intelligence (AI)-DTW method, which was then tested on 40 patients with general LV morphology, 20 with dilated cardiomyopathy (DCMP), and 20 with transthyretin-associated cardiac amyloidosis (ATTR-CA), 20 with severe aortic stenosis (AS), and 20 with severe mitral regurgitation (MR). Experiments-1 and -2 revealed that the DTW method is consistent with dedicated software. In experiment-3, the AI-DTW strain method showed comparable results for general LV morphology (bias - 0.137 ± 0.398%), DCMP (- 0.397 ± 0.607%), ATTR-CA (0.095 ± 0.581%), AS (0.334 ± 0.358%), and MR (0.237 ± 0.490%). Moreover, the strain curves showed a high correlation in their characteristics, with R-squared values of 0.8879-0.9452 for those LV morphology in experiment-3. Measuring LVGLS through dynamic warping of segmentation contour is a feasible method compared to traditional tracking techniques. This approach has the potential to decrease the need for manual demarcation and make LVGLS measurements more efficient and user-friendly for daily practice.

Learning-Based Regularization for Cardiac Strain Analysis via Domain Adaptation.

Reliable motion estimation and strain analysis using 3D+ time echocardiography (4DE) for localization and characterization of myocardial injury is valuable for early detection and targeted interventions. However, motion estimation is difficult due to the low-SNR that stems from the inherent image properties of 4DE, and intelligent regularization is critical for producing reliable motion estimates. In this work, we incorporated the notion of domain adaptation into a supervised neural network regularization framework. We first propose a semi-supervised Multi-Layered Perceptron (MLP) network with biomechanical constraints for learning a latent representation that is shown to have more physiologically plausible displacements. We extended this framework to include a supervised loss term on synthetic data and showed the effects of biomechanical constraints on the network's ability for domain adaptation. We validated the semi-supervised regularization method on in vivo data with implanted sonomicrometers. Finally, we showed the ability of our semi-supervised learning regularization approach to identify infarct regions using estimated regional strain maps with good agreement to manually traced infarct regions from postmortem excised hearts.

Laser-Generated Leaky Acoustic Wave Imaging for Interventional Guidewire Guidance.

Ultrasound (US) is widely used to visualize both tissue and the positions of surgical instruments in real time during surgery. Previously we proposed a new method to exploit US imaging and laser-generated leaky acoustic waves (LAWs) for needle visualization. Although successful, that method only detects the position of a needle tip, with the location of the entire needle deduced from knowing that the needle is straight. The purpose of the current study was to develop a beamforming-based method for the direct visualization of objects. The approach can be applied to objects with arbitrary shapes, such as the guidewires that are commonly used in interventional guidance. With this method, illumination by a short laser pulse generates photoacoustic waves at the top of the guidewire that propagate down its metal surface. These waves then leak into the surrounding tissue, which can be detected by a US array transducer. The time of flight consists of two parts: 1) the propagation time of the guided waves on the guidewire and 2) the propagation time of the US that leaks into the tissue. In principle, an image of the guidewire can be formed based on array beamforming by taking the propagation time on the metal into consideration. Furthermore, we introduced directional filtering and a matched filter to compress the dispersion signal associated with long propagation times. The results showed that guidewires could be detected at depths of at least 70 mm. The maximum detectable angle was 56.3°. LAW imaging with a 1268-mm-long guidewire was also demonstrated. The proposed method has considerable potential in new clinical applications.

Real-time interleaved spectroscopic photoacoustic and ultrasound (PAUS) scanning with simultaneous fluence compensation and motion correction.

For over two decades photoacoustic imaging has been tested clinically, but successful human trials have been limited. To enable quantitative clinical spectroscopy, the fundamental issues of wavelength-dependent fluence variations and inter-wavelength motion must be overcome. Here we propose a real-time, spectroscopic photoacoustic/ultrasound (PAUS) imaging approach using a compact, 1-kHz rate wavelength-tunable laser. Instead of illuminating tissue over a large area, the fiber-optic delivery system surrounding an US array sequentially scans a narrow laser beam, with partial PA image reconstruction for each laser pulse. The final image is then formed by coherently summing partial images. This scheme enables (i) automatic compensation for wavelength-dependent fluence variations in spectroscopic PA imaging and (ii) motion correction of spectroscopic PA frames using US speckle tracking in real-time systems. The 50-Hz video rate PAUS system is demonstrated in vivo using a murine model of labelled drug delivery.

Spatially localized sono-photoacoutic activation of phase-change contrast agents.

Sono-photoacoustic (SPA) activation lowers the threshold of phase-change contrast agents by timing a laser shot to coincide with the arrival of an acoustic wave at a region of interest. The combination of photothermal heating from optical absorption and negative pressure from the acoustic wave greatly reduces the droplet's combined vaporization threshold compared to using laser energy or acoustic energy alone. In previous studies, SPA imaging used a broadly illuminated optical pulse combined with plane wave acoustic pulses transmitted from a linear ultrasound array. Acoustic plane waves cover a wide lateral field of view, enabling direct visualization of the contrast agent distribution. In contrast, we demonstrate here that localized SPA activation is possible using electronically steered/focused ultrasound pulses. The focused SPA activation region is defined axially by the number of cycles in the acoustic pulse and laterally by the acoustic beam width. By reducing the spot size and enabling rapid electronic steering, complex activation patterns are possible, which may be particularly useful in therapeutic applications.

Correction of wavelength-dependent laser fluence in swept-beam spectroscopic photoacoustic imaging with a hand-held probe.

Recently, we demonstrated an integrated photoacoustic (PA) and ultrasound (PAUS) system using a kHz-rate wavelength-tunable laser and a swept-beam delivery approach. It irradiates a medium using a narrow laser beam swept at high repetition rate (∼1 kHz) over the desired imaging area, in contrast to the conventional PA approach using broad-beam illumination at a low repetition rate (10-50 Hz). Here, we present a method to correct the wavelength-dependent fluence distribution and demonstrate its performance in phantom studies using a conventional limited view/bandwidth hand-held US probe. We adopted analytic fluence models, extending diffusion theory for the case of a pencil beam obliquely incident on an optically homogenous turbid medium, and developed a robust method to estimate fluence attenuation in the medium using PA measurements acquired from multiple fiber-irradiation positions swept at a kHz rate. We conducted comprehensive simulation tests and phantom studies using well-known contrast-agents to validate the reliability of the fluence model and its spectral corrections.

Deep-Learning Image Reconstruction for Real-Time Photoacoustic System.

Recent advances in photoacoustic (PA) imaging have enabled detailed images of microvascular structure and quantitative measurement of blood oxygenation or perfusion. Standard reconstruction methods for PA imaging are based on solving an inverse problem using appropriate signal and system models. For handheld scanners, however, the ill-posed conditions of limited detection view and bandwidth yield low image contrast and severe structure loss in most instances. In this paper, we propose a practical reconstruction method based on a deep convolutional neural network (CNN) to overcome those problems. It is designed for real-time clinical applications and trained by large-scale synthetic data mimicking typical microvessel networks. Experimental results using synthetic and real datasets confirm that the deep-learning approach provides superior reconstructions compared to conventional methods.

Flow network tracking for spatiotemporal and periodic point matching: Applied to cardiac motion analysis.

The accurate quantification of left ventricular (LV) deformation/strain shows significant promise for quantitatively assessing cardiac function for use in diagnosis and therapy planning. However, accurate estimation of the displacement of myocardial tissue and hence LV strain has been challenging due to a variety of issues, including those related to deriving tracking tokens from images and following tissue locations over the entire cardiac cycle. In this work, we propose a point matching scheme where correspondences are modeled as flow through a graphical network. Myocardial surface points are set up as nodes in the network and edges define neighborhood relationships temporally. The novelty lies in the constraints that are imposed on the matching scheme, which render the correspondences one-to-one through the entire cardiac cycle, and not just two consecutive frames. The constraints also encourage motion to be cyclic, which an important characteristic of LV motion. We validate our method by applying it to the estimation of quantitative LV displacement and strain estimation using 8 synthetic and 8 open-chested canine 4D echocardiographic image sequences, the latter with sonomicrometric crystals implanted on the LV wall. We were able to achieve excellent tracking accuracy on the synthetic dataset and observed a good correlation with crystal-based strains on the in-vivo data.

Ultrasound-based formation of nano-Pickering emulsions investigated via in-situ SAXS.

Sonication is one of the most commonly used methods to synthesize Pickering emulsions. Yet, the process of emulsion sonication is rarely characterized in detail and acoustic conditions are largely determined by experimenter's personal experience. In this study, the role of sonication in the formation of Pickering emulsions from amphiphilic gold nanoparticles was investigated using a new sample environment combining ultrasound delivery with ultra-small-angle X-ray scattering (USAXS) measurements. The detection of acoustic cavitation and the simultaneous analysis of structural data via USAXS demonstrated direct correlation between Pickering emulsion formation and cavitation events. There was no evidence of spontaneous adsorption of particles onto the oil-water interface without ultrasound, which suggests the presence of a stabilizing force. Acoustically detected cavitation events could originate in the bulk solvent and/or inside the emulsion droplets. These events helped overcome energy barriers to induce particle adsorption.

Efficient Two-Pass 3-D Speckle Tracking for Ultrasound Imaging.

Speckle tracking based on block matching is the most common method for multi-dimensional motion estimation in ultrasound elasticity imaging. Extension of two-dimensional (2-D) methods to three dimensions (3-D) has been problematic because of the large computational load of 3-D tracking, as well as performance issues related to the low frame (volume) rates of 3-D images. To address both of these problems, we have developed an efficient two-pass tracking method suited to cardiac elasticity imaging. PatchMatch, originally developed for image editing, has been adapted for ultrasound to provide first-pass displacement estimates. Second-pass estimation uses conventional block matching within a much smaller search region. 3-D displacements are then obtained using correlation filtering previously shown to be effective against speckle decorrelation. Both simulated and in vivo canine cardiac results demonstrate that the proposed two-pass method reduces computational cost compared to conventional 3-D exhaustive search by a factor of 10. Moreover, it outperforms one-pass tracking by a factor of about 3 in terms of root-mean-square error relative to available ground-truth displacements.

ECG triggering and gating for ultrasonic small animal imaging.

Echocardiography (ECG) is routinely used in the clinical diagnosis of cardiac function. The anatomy of the mouse is similar to that of the human, and thus murine ECG has become an effective tool for the assessment of small animal models of human cardiac diseases. Unfortunately, clinical ultrasonic imaging systems are not suitable for murine cardiac imaging due to their limited spatial and temporal resolutions. Murine ECG requires a spatial resolution better than 100 pim, which mandates the use of high-frequency, ultrasonic imaging (i.e., >20 MHz). High-frequency transducer arrays currently are not available, and so such systems use the mechanical scanning of a single-element transducer for which the frame rate is insufficient for directly monitoring the rapid beating of a mouse heart, and thus retrospective image reconstruction is necessary. This paper presents a high-frequency, ultrasonic imaging system for murine cardiac imaging. Two scanning methods have been developed. One is based on ECG triggering and is called the block scanning mode, in which the murine cardiac images from the isovolumic contraction and isovolumic relaxation phases are retrospectively reconstructed within a relatively short data acquisition time using the ECG R-wave as the trigger to the imaging system. The other method is the line scanning mode based on ECG gating, in which both ECG and ultrasound scan lines are continuously acquired over a longer time, enabling images during the entire cardiac cycle to be obtained. It is demonstrated here that the effective frame rate is determined by the pulse repetition frequency and can be up to 2 kHz in the presented system.

Vector-velocity estimation in swept-scan using a K-space approach.

The swept-scan technique (i.e., continuously moving a single-crystal transducer during pulse-echo data acquisition) is used in high-frequency, ultrasonic flow imaging. Relative to the conventional step-scan technique, swept scanning improves the rate of data acquisition and enables near-real-time, high-frequency color flow mapping. However, the continuous transducer movement may have non-negligible effects on accuracy of velocity estimation. This paper introduces a spatial frequency domain (i.e., k-space) approach that quantifies the effects of both lateral and axial motions in a swept scan. It is shown that the k-space representation is equivalent to a Doppler-radio frequency (RF) frequency domain representation, and that transducer movement in the swept-scan technique results in a change in Doppler bandwidth. In addition, a vector velocity estimator is developed based on the proposed k-space approach. Both simulations and flow-phantom experiments were performed to evaluate the performance of the proposed vector velocity estimator. A 45-MHz transducer was scanned at 20 mm/s. The Doppler angle ranged from 29 degrees to 90 degrees, and the flow velocities ranged from 15 to 30 mm/s. The results show that the proposed k-space vector velocity estimator exhibited a mean error of 2.6 degrees for flow-direction estimation, with the standard deviation ranging from 2.2 degrees to 8.2 degrees. In comparison, for the conventional spectral-broadening-based vector velocity estimator ignoring the swept-scan effect, the mean error became 15 degrees and the standard deviations were from 2.7 degrees to 6.6 degrees.

Effects of swept scanning on velocity estimation.

The swept-scan technique adopted in high-frequency ultrasound involves mechanically scanning a single-element transducer to acquire image data. Unlike conventional step scanning, where the image data are acquired at discrete positions, the swept-scan technique acquires the image data while the transducer is continuously moving. Such a scanning method is particularly advantageous for Doppler flow estimation because its frame rate is higher than that for the step-scan technique. However, the effects of the transducer motion on the accuracy of velocity estimation have not been studied comprehensively. This study employed a k-space approach to experimentally investigate the effects of swept scanning on both conventional Doppler axial velocity estimation and spectral-broadening-based lateral velocity estimation using a 45-MHz transducer. The results indicate that such effects must be corrected in order to obtain an accurate estimation of flow velocities.

ECG Gated Ultrasonic Small Animal Imaging.

Echocardiography is a routine clinical procedure to diagnose cardiac functions. The organic structure of the mouse is similar to that of human so that murine echocardiography has potentially become an effective tool for the assessment of human cardiovascular disease. However, clinical ultrasonic imaging systems are not suitable for murine cardiac imaging due to its limited spatial and temporal resolution. Thus, high frequency ultrasonic imaging (≥ 20 MHz) is necessary in order to provide spatial resolution at the order of 100 μm. Furthermore, due to the lack of transducer arrays at such a high frequency, single-element transducer with mechanical scanning is typically used. Thus the frame rate is insufficient for imaging the quick motion of the mouse. In this paper, a high frequency ultrasonic imaging system with electrocardiography gating is built in order to provide both high spatial resolution and high temporal effecting resolution. The system utilizes the R-wave trigger signal from murine electrocardiography. Image data are acquired in either the block scanning mode or the line scanning mode. In block scanning, murine cardiac images in systole and diastole can be retrospectively reconstructed with a short data acquisition time. In line scanning, on the other hand, images during the entire cardiac cycle can be obtained. It is demonstrated that the effective frame rate can be up to 2 kHz, which is only limited by the pulse repetition rate of the system.