Atherosclerosis is the major cause of cardiovascular diseases (CVDs). Intravascular ultrasound (IVUS) is a common imaging modality for diagnosing CVDs. However, an efficient analyzer for IVUS image segmentation is required for assisting cardiologists. In this study, an end-to-end deep-learning convolutional neural network was developed for automatically detecting media–adventitia borders, luminal regions, and calcified plaque in IVUS images. A total of 713 grayscale IVUS images from 18 patients were used as training data for the proposed deep-learning model. The model is constructed using the three modified U-Nets and combined with the concept of cascaded networks to prevent errors in the detection of calcification owing to the interference of pixels outside the plaque regions. Three loss functions (Dice, Tversky, and focal loss) with various characteristics were tested to determine the best setting for the proposed model. The efficacy of the deep-learning model was evaluated by analyzing precision–recall curve. The average precision (AP), Dice score coefficient, precision, sensitivity, and specificity of the predicted and ground truth results were then compared. All training processes were validated using leave-one-subject-out cross-validation. The experimental results showed that the proposed deep-learning model exhibits high performance in segmenting the media–adventitia layers and luminal regions for all loss functions, with all tested metrics being higher than 0.90. For locating calcified tissues, the best result was obtained when the focal loss function was applied to the proposed model, with an AP of 0.73; however, the prediction efficacy was affected by the proportion of calcified tissues within the plaque region when the focal loss function was employed. Compared with commercial software, the proposed method exhibited high accuracy in segmenting IVUS images in some special cases, such as when shadow artifacts or side vessels surrounded the target vessel.

Zebrafish has been recently considered an ideal vertebrate for studying developmental biology, genetics, particularly for modeling tumorigenesis, angiogenesis, and regeneration in vivo. However, when a zebrafish matures completely, its body loses transparency, thus making conventional optical imaging techniques difficult for imaging internal anatomy and vasculature. Acoustic wave penetration outperforms optical methods, high-frequency (>30 MHz) ultrasound (HFUS) was consequently an alternative imaging modality for adult zebrafish imaging, particularly for echocardiography However, visualizing peripheral vessels in a zebrafish by using conventional HFUS is still difficult. In the present study, high-frequency micro-Doppler imaging (HFμDI) based on ultrafast ultrasound imaging was proposed for zebrafish dorsal vascular mapping in vivo. HFμDI uses a 40-MHz ultrasound transducer, which is an ultrafast ultrasound imaging technology with the highest frequency available currently. Blood flow signals were extracted using an eigen-based clutter filter with different settings. Experiments were performed on an 8-month-old wild-type AB-line adult zebrafish. Blood vessels, including intersegmental vessels, parachordal vessel, dorsal longitudinal anastomotic vessel, and dorsal aorta, from the dorsal side of the zebrafish were clearly observed in two-dimensional (2-D) and 3-D HFμDI. The maximum image depth of HFμDI and the minimal diameter of vessel can be detected were 4 mm and 36 μm, respectively; they were determined without any use of microbubbles. The maximum flow velocity range was approximately 3-4 mm/s on the dorsal vessels of the adult zebrafish. Compared with conventional ultrasound Doppler imaging, HFμDI exhibited superior small vessel imaging.

Characterizing the viscoelastic properties of thin-layer tissues with micro-level thickness has long remained challenging. Recently, several micro-elastography techniques have been developed to improve the spatial resolution. However, most of these techniques have not considered the medium boundary conditions when evaluating the viscoelastic properties of thin-layer tissues such as arteries and corneas; this might lead to estimation bias or errors. This paper aims to integrate the Lamb wave model with our previously developed ultrasonic micro-elastography imaging system for obtaining accurate viscoelastic properties in thin-layer tissues. A 4.5-MHz ring transducer was used to generate an acoustic radiation force for inducing tissue displacements to produce guided wave, and the wave propagation was detected using a confocally aligned 40-MHz needle transducer. The phase velocity and attenuation were obtained from k-space by both the impulse and the harmonic methods. The measured phase velocity was fit using the Lamb wave model with the Kelvin-Voigt model. Phantom experiments were conducted using 7% and 12% gelatin and 1.5% agar phantoms with different thicknesses (2, 3, and 4 mm). Biological experiments were performed on porcine cornea and rabbit carotid artery ex vivo . Thin-layer phantoms with different thicknesses were confirmed to have the same elasticity; this was consistent with the estimates of bulk phantoms from mechanical tests and the shear wave rheological model. The trend of the measured attenuations was also confirmed with the viscosity results obtained using the Lamb wave model. Through the impulse and harmonic methods, the shear viscoelasticity values were estimated to be 8.2 kPa for 0.9 Pa·s and 9.6 kPa for 0.8 Pa·s in the cornea and 27.9 kPa for 0.1 Pa·s and 26.5 kPa for 0.1 Pa·s in the artery.

Ultrafast ultrasound imaging based on plane wave (PW) compounding has been proposed for use in various clinical and preclinical applications, including shear wave imaging and super resolution blood flow imaging. Because the image quality afforded by PW imaging is highly dependent on the number of PW angles used for compounding, a tradeoff between image quality and frame rate occurs. In the present study, a convolutional neural network (CNN) beamformer based on a combination of the GoogLeNet and U-Net architectures was developed to replace the conventional delay-and-sum (DAS) algorithm to obtain high-quality images at a high frame rate. RF channel data are used as the inputs for the CNN beamformers. The outputs are in-phase and quadrature data. Simulations and phantom experiments revealed that the images predicted by the CNN beamformers had higher resolution and contrast than those predicted by conventional single-angle PW imaging with the DAS approach. In in vivo studies, the contrast-to-noise ratios (CNRs) of carotid artery images predicted by the CNN beamformers using three or five PWs as ground truths were approximately 12 dB in the transverse view, considerably higher than the CNR obtained using the DAS beamformer (3.9 dB). Most tissue speckle information was retained in the in vivo images produced by the CNN beamformers. In conclusion, only a single PW at 0° was fired, but the quality of the output image was proximal to that of an image generated using three or five PW angles. In other words, the quality–frame rate tradeoff of coherence compounding could be mitigated through the use of the proposed CNN for beamforming.