Delineation of the cardiac structures from 2D echocardiographic images is a common clinical task to establish a diagnosis. Over the past decades, the automation of this task has been the subject of intense research. In this paper, we evaluate how far the state-of-the-art encoder-decoder deep convolutional neural network methods can go at assessing 2D echocardiographic images, i.e., segmenting cardiac structures and estimating clinical indices, on a dataset, especially, designed to answer this objective. We, therefore, introduce the cardiac acquisitions for multi-structure ultrasound segmentation dataset, the largest publicly-available and fully-annotated dataset for the purpose of echocardiographic assessment. The dataset contains two and four-chamber acquisitions from 500 patients with reference measurements from one cardiologist on the full dataset and from three cardiologists on a fold of 50 patients. Results show that encoder-decoder-based architectures outperform state-of-the-art non-deep learning methods and faithfully reproduce the expert analysis for the end-diastolic and end-systolic left ventricular volumes, with a mean correlation of 0.95 and an absolute mean error of 9.5 ml. Concerning the ejection fraction of the left ventricle, results are more contrasted with a mean correlation coefficient of 0.80 and an absolute mean error of 5.6%. Although these results are below the inter-observer scores, they remain slightly worse than the intra-observer's ones. Based on this observation, areas for improvement are defined, which open the door for accurate and fully-automatic analysis of 2D echocardiographic images.

Left ventricular myocardial stiffness could offer superior quantification of cardiac systolic and diastolic function when compared to the current diagnostic tools. Shear wave elastography in combination with acoustic radiation force has been widely proposed to noninvasively assess tissue stiffness. Interestingly, shear waves can also result from intrinsic cardiac mechanical events (e.g., closure of valves) without the need for external excitation. However, it remains unknown whether these natural shear waves always occur, how reproducible they can be detected and what the normal range of shear wave propagation speed is. The present study, therefore, aimed at establishing the feasibility of detecting shear waves created after mitral valve closure (MVC) and aortic valve closure (AVC), the variability of the measurements, and at reporting the normal values of propagation velocity. Hereto, a group of 30 healthy volunteers was scanned with high-frame rate imaging (>1000 Hz) using an experimental ultrasound system transmitting a diverging wave sequence. Tissue Doppler velocity and acceleration were used to create septal color M-modes, on which the shear waves were tracked and their velocities measured. Overall, the methodology was capable of detecting the transient vibrations that spread throughout the intraventricular septum in response to the closure of the cardiac valves in 92% of the recordings. Reference velocities of 3.2±0.6 m/s at MVC and 3.5±0.6 m/s at AVC were obtained. Moreover, in order to show the diagnostic potential of this approach, two patients (one with cardiac amyloidosis and one undergoing a dobutamine stress echocardiography) were scanned with the same protocol and showed markedly higher propagation speeds: the former presented velocities of 6.6 and 5.6 m/s; the latter revealed normal propagation velocities at baseline, and largely increased during the dobutamine infusion (>15 m/s). Both cases showed values consistent with the expected changes in stiffness and cardiac loading conditions.

Cardiac volume/function assessment remains a critical step in daily cardiology and 3D ultrasound plays an increasingly important role. Fully automatic left ventricular segmentation is, however, a challenging task due to the artifacts and low contrast-to-noise ratio of ultrasound imaging. In the present work, a fast and fully automatic framework for full cycle endocardial left ventricle segmentation is proposed. This approach couples the advantages of the B-spline explicit active surfaces framework, a purely image information approach, to those of statistical shape models to give prior information about the expected shape for an accurate segmentation. The segmentation is propagated throughout the heart cycle using a localized anatomical affine optical flow. It is shown that this approach not only outperforms other state-of-the-art methods in terms of distance metrics with mean average distances of 1.81±0.59 mm and 1.98±0.66 mm at end-diastole and end-systole respectively but is computationally efficient (in average 11 seconds per 4D image) and fully automatic.

Aims: To determine myocardial stiffness by means of measuring the velocity of naturally occurring myocardial shear waves (SWs) at mitral valve closure (MVC) and investigate their changes with myocardial remodelling in patients with hypertensive heart disease. Methods and results: Thirty-three treated arterial hypertension (HT) patients with hypertrophic left ventricular (LV) remodelling (59 ± 14 years, 55% male) and 26 aged matched healthy controls (55±15 years, 77% male) were included. HT patients were further divided into a concentric remodelling (HT1) group (13 patients) and a concentric hypertrophy (HT2) group (20 patients). LV parasternal long-axis views were acquired with an experimental ultrasound scanner at 1266 ± 317 frames per seconds. The SW velocity induced by MVC was measured from myocardial acceleration maps. SW velocities differed significantly between HT patients and controls (5.83 ± 1.20 m/s vs. 4.04 ± 0.96 m/s; P < 0.001). In addition, the HT2 group had the highest SW velocities (P < 0.001), whereas values between controls and the HT1 group were comparable (P = 0.075). Significant positive correlations were found between SW velocity and LV remodelling (interventricular septum thickness: r = 0.786, P < 0.001; LV mass index: r = 0.761, P < 0.001). SW velocity normalized for wall stress indicated that myocardial stiffness in the HT2 group was twice as high as in controls (P < 0.001), whereas values of the HT1 group overlapped with the controls (P = 1.00). Conclusions: SW velocity as measure of myocardial stiffness is higher in HT patients compared with healthy controls, particularly in advanced hypertensive heart disease. Patients with concentric remodelling have still normal myocardial properties whereas patients with concentric hypertrophy show significant stiffening.