Ultrasound Shear Wave Elastography in Cardiology.

The advent of high-frame rate imaging in ultrasound allowed the development of shear wave elastography as a noninvasive alternative for myocardial stiffness assessment. It measures mechanical waves propagating along the cardiac wall with speeds that are related to stiffness. The use of cardiac shear wave elastography in clinical studies is increasing, but a proper understanding of the different factors that affect wave propagation is required to correctly interpret results because of the heart's thin-walled geometry and intricate material properties. The aims of this review are to give an overview of the general concepts in cardiac shear wave elastography and to discuss in depth the effects of age, hemodynamic loading, cardiac morphology, fiber architecture, contractility, viscoelasticity, and system-dependent factors on the measurements, with a focus on clinical application. It also describes how these factors should be considered during acquisition, analysis, and reporting to ensure an accurate, robust, and reproducible measurement of the shear wave.

Continuous shear wave measurements for dynamic cardiac stiffness evaluation in pigs.

Ultrasound-based shear wave elastography is a promising technique to non-invasively assess the dynamic stiffness variations of the heart. The technique is based on tracking the propagation of acoustically induced shear waves in the myocardium of which the propagation speed is linked to tissue stiffness. This measurement is repeated multiple times across the cardiac cycle to assess the natural variations in wave propagation speed. The interpretation of these measurements remains however complex, as factors such as loading and contractility affect wave propagation. We therefore applied transthoracic shear wave elastography in 13 pigs to investigate the dependencies of wave speed on pressure-volume derived indices of loading, myocardial stiffness, and contractility, while altering loading and inducing myocardial ischemia/reperfusion injury. Our results show that diastolic wave speed correlates to a pressure-volume derived index of operational myocardial stiffness (R = 0.75, p < 0.001), suggesting that both loading and intrinsic properties can affect diastolic wave speed. Additionally, the wave speed ratio, i.e. the ratio of systolic and diastolic speed, correlates to a pressure-volume derived index of contractility, i.e. preload-recruitable stroke work (R = 0.67, p < 0.001). Measuring wave speed ratio might thus provide a non-invasive index of contractility during ischemia/reperfusion injury.

Deep neural network-based clustering of deformation curves reveals novel disease features in PLN pathogenic variant carriers.

Echocardiographic deformation curves provide detailed information on myocardial function. Deep neural networks (DNNs) may enable automated detection of disease features in deformation curves, and improve the clinical assessment of these curves. We aimed to investigate whether an explainable DNN-based pipeline can be used to detect and visualize disease features in echocardiographic deformation curves of phospholamban (PLN) p.Arg14del variant carriers. A DNN was trained to discriminate PLN variant carriers (n = 278) from control subjects (n = 621) using raw deformation curves obtained by 2D-speckle tracking in the longitudinal axis. A visualization technique was used to identify the parts of these curves that were used by the DNN for classification. The PLN variant carriers were clustered according to the output of the visualization technique. The DNN showed excellent discriminatory performance (C-statistic 0.93 [95% CI 0.87-0.97]). We identified four clusters with PLN-associated disease features in the deformation curves. Two clusters showed previously described features: apical post-systolic shortening and reduced systolic strain. The two other clusters revealed novel features, both reflecting delayed relaxation. Additionally, a fifth cluster was identified containing variant carriers without disease features in the deformation curves, who were classified as controls by the DNN. This latter cluster had a very benign disease course regarding development of ventricular arrhythmias. Applying an explainable DNN-based pipeline to myocardial deformation curves enables automated detection and visualization of disease features. In PLN variant carriers, we discovered novel disease features which may improve individual risk stratification. Applying this approach to other diseases will further expand our knowledge on disease-specific deformation patterns. Overview of the deep neural network-based pipeline for feature detection in myocardial deformation curves. Firstly, phospholamban (PLN) p.Arg14del variant carriers and controls were selected and a deep neural network (DNN) was trained to detect the PLN variant carriers. Subsequently, a clustering-based approach was performed on the attention maps of the DNN, which revealed 4 distinct phenotypes of PLN variant carriers with different prognoses. Moreover, a cluster without features and a benign prognosis was detected.

Correcting bias in cardiac geometries derived from multimodal images using spatiotemporal mapping.

Cardiovascular imaging studies provide a multitude of structural and functional data to better understand disease mechanisms. While pooling data across studies enables more powerful and broader applications, performing quantitative comparisons across datasets with varying acquisition or analysis methods is problematic due to inherent measurement biases specific to each protocol. We show how dynamic time warping and partial least squares regression can be applied to effectively map between left ventricular geometries derived from different imaging modalities and analysis protocols to account for such differences. To demonstrate this method, paired real-time 3D echocardiography (3DE) and cardiac magnetic resonance (CMR) sequences from 138 subjects were used to construct a mapping function between the two modalities to correct for biases in left ventricular clinical cardiac indices, as well as regional shape. Leave-one-out cross-validation revealed a significant reduction in mean bias, narrower limits of agreement, and higher intraclass correlation coefficients for all functional indices between CMR and 3DE geometries after spatiotemporal mapping. Meanwhile, average root mean squared errors between surface coordinates of 3DE and CMR geometries across the cardiac cycle decreased from 7 ± 1 to 4 ± 1 mm for the total study population. Our generalised method for mapping between time-varying cardiac geometries obtained using different acquisition and analysis protocols enables the pooling of data between modalities and the potential for smaller studies to leverage large population databases for quantitative comparisons.

Analytic prediction of droplet vaporization events to estimate the precision of ultrasound-based proton range verification.

BACKGROUND: The safety and efficacy of proton therapy is currently hampered by range uncertainties. The combination of ultrasound imaging with injectable radiation-sensitive superheated nanodroplets was recently proposed for in vivo range verification. The proton range can be estimated from the distribution of nanodroplet vaporization events, which is stochastically related to the stopping distribution of protons, as nanodroplets are vaporized by protons reaching their maximal LET at the end of their range. PURPOSE: Here, we aim to estimate the range estimation precision of this technique. As for any stochastic measurement, the precision will increase with the sample size, that is, the number of detected vaporizations. Thus, we first develop and validate a model to predict the number of vaporizations, which is then applied to estimate the range verification precision for a set of conditions (droplet size, droplet concentration, and proton beam parameters). METHODS: Starting from the thermal spike theory, we derived a model that predicts the expected number of droplet vaporizations in an irradiated sample as a function of the droplet size, concentration, and number of protons. The model was validated by irradiating phantoms consisting of size-sorted perfluorobutane droplets dispersed in an aqueous matrix. The number of protons was counted with an ionization chamber, and the droplet vaporizations were recorded and counted individually using high frame rate ultrasound imaging. After validation, the range estimate precision was determined for different conditions using a Monte Carlo algorithm. RESULTS: A good agreement between theory and experiments was observed for the number of vaporizations, especially for large (5.8 ± 2.2 µm) and medium (3.5 ± 1.1 µm) sized droplets. The number of events was lower than expected in phantoms with small droplets (2.0 ± 0.7 µm), but still within the same order of magnitude. The inter-phantom variability was considerably larger (up to 30x) than predicted by the model. The validated model was then combined with Monte Carlo simulations, which predicted a theoretical range retrieval precision improving with the square-root of the number of vaporizations, and degrading at high beam energies due to range straggling. For single pencil beams with energies between 70 and 240 MeV, a range verification precision below 1% of the range required perfluorocarbon concentrations in the order of 0.3-2.4 µM. CONCLUSION: We proposed and experimentally validated a model to provide a quick estimate of the number of vaporizations for a given set of conditions (droplet size, droplet concentration, and proton beam parameters). From this model, promising range verification performances were predicted for realistic perfluorocarbon concentrations. These findings are an incentive to move towards preclinical studies, which are critical to assess the achievable droplet distribution in and around the tumor, and hence the in vivo range verification precision.

Transmural Wave Speed Gradient May Distinguish Intrinsic Myocardial Stiffening From Preload-Induced Changes in Operational Stiffness in Shear Wave Elastography.

BACKGROUND: Shear wave elastography (SWE) is a promising technique to non-invasively assess myocardial stiffness based on the propagation speed of mechanical waves. However, a high wave propagation speed can either be attributed to an elevated intrinsic myocardial stiffness or to a preload-induced increase in operational stiffness. OBJECTIVE: Our objective was to find a way to discriminate intrinsic myocardial stiffening from stiffening caused by an increased pressure in SWE. METHODS: We used the finite element method to study the shear wave propagation patterns when stiffness and/or pressure is elevated, compared to normal stiffness and pressure. Numerical findings were verified in a few human subjects. RESULTS: The transmural wave speed gradient was able to distinguish changes in intrinsic stiffness from those induced by differing hemodynamic load (a speed of ±3.2 m/s in parasternal short-axis (PSAX) view was associated with a wave speed gradient of -0.17 ± 0.15 m/s/mm when pressure was elevated compared to 0.04 ± 0.05 m/s/mm when stiffness was elevated). The gradient however decreased when stiffness increased (decrease with a factor 3 in PSAX when stiffness doubled at 20 mmHg). The human data analysis confirmed the presence of a wave speed gradient in a patient with elevated ventricular pressure. CONCLUSION: Cardiac SWE modeling is a useful tool to gain additional insights into the complex wave physics and to guide post-processing. The transmural differences in wave speed may help to distinguish loading-induced stiffening from intrinsic stiffness changes. SIGNIFICANCE: The transmural wave speed gradient has potential as a new diagnostic parameter for future clinical studies.

Use of 3D anatomical models in mock circulatory loops for cardiac medical device testing.

INTRODUCTION: Mock circulatory loops (MCLs) are mechanical representations of the cardiovascular system largely used to test the hemodynamic performance of cardiovascular medical devices (MD). Thanks to 3 dimensional (3D) printing technologies, MCLs can nowadays also incorporate anatomical models so to offer enhanced testing capabilities. The aim of this review is to provide an overview on MCLs and to discuss the recent developments of 3D anatomical models for cardiovascular MD testing. METHODS: The review first analyses the different techniques to develop 3D anatomical models, in both rigid and compliant materials. In the second section, the state of the art of MCLs with 3D models is discussed, along with the testing of different MDs: implantable blood pumps, heart valves, and imaging techniques. For each class of MD, the MCL is analyzed in terms of: the cardiovascular model embedded, the 3D model implemented (the anatomy represented, the material used, and the activation method), and the testing applications. DISCUSSIONS AND CONCLUSIONS: MCLs serve the purpose of testing cardiovascular MDs in different (patho-)physiological scenarios. The addition of 3D anatomical models enables more realistic connections of the MD with the implantation site and enhances the testing capabilities of the MCL. Current attempts focus on the development of personalized MCLs to test MDs in patient-specific hemodynamic and anatomical scenarios. The main limitation of MCLs is the impossibility to assess the impact of a MD in the long-term and at a biological level, for which animal experiments are still needed.

Impact of Loading and Myocardial Mechanical Properties on Natural Shear Waves: Comparison to Pressure-Volume Loops.

BACKGROUND: Shear wave elastography (SWE) has been proposed as a novel noninvasive method for the assessment of myocardial stiffness, a relevant determinant of diastolic function. It is based on tracking the propagation of shear waves, induced, for instance, by mitral valve closure (MVC), in the myocardium. The speed of propagation is directly related to myocardial stiffness, which is defined by the local slope of the nonlinear stress-strain relation. Therefore, the operating myocardial stiffness can be altered by both changes in loading and myocardial mechanical properties. OBJECTIVES: This study sought to evaluate the capability of SWE to quantify myocardial stiffness changes in vivo by varying loading and myocardial tissue properties and to compare SWE against pressure-volume loop analysis, a gold standard reference method. METHODS: In 15 pigs, conventional and high-frame rate echocardiographic data sets were acquired simultaneously with pressure-volume loop data after acutely changing preload and afterload and after inducting an ischemia/reperfusion (I/R) injury. RESULTS: Shear wave speed after MVC significantly increased by augmenting preload and afterload (3.2 ± 0.8 m/s vs 4.6 ± 1.2 m/s and 4.6 ± 1.0 m/s, respectively; P = 0.001). Preload reduction had no significant effect on shear wave speed compared to baseline (P = 0.118). I/R injury resulted in significantly higher shear wave speed after MVC (6.1 ± 1.2 m/s; P < 0.001). Shear wave speed after MVC had a strong correlation with the chamber stiffness constant ß (r = 0.63; P < 0.001) and operating chamber stiffness dP/dV before induction of an I/R injury (r = 0.78; P < 0.001) and after (r = 0.83; P < 0.001). CONCLUSIONS: Shear wave speed after MVC was influenced by both acute changes in loading and myocardial mechanical properties, reflecting changes in operating myocardial stiffness, and was strongly related to chamber stiffness, invasively derived by pressure-volume loop analysis. SWE provides a novel noninvasive method for the assessment of left ventricular myocardial properties.

Comparing Myocardial Shear Wave Propagation Velocity Estimation Methods Based on Tissue Displacement, Velocity and Acceleration Data.

Shear wave elastography (SWE) is a promising technique used to assess cardiac function through the evaluation of cardiac stiffness non-invasively. However, in the literature, SWE varies in terms of tissue motion data (displacement, velocity or acceleration); method used to characterize mechanical wave propagation (time domain [TD] vs. frequency domain [FD]); and the metric reported (wave speed [WS], shear or Young's modulus). This variety of reported methodologies complicates comparison of reported findings and sheds doubt on which methodology better approximates the true myocardial properties. We therefore conducted a simulation study to investigate the accuracy of various SWE data analysis approaches while varying cardiac geometry and stiffness. Lower WS values were obtained by the TD method compared with the FD method. Acceleration-based WS estimates in the TD were systematically larger than those based on velocity (∼10% difference). These observations were confirmed by TD analysis of 32 in vivo SWE mechanical wave measurements. In vivo data quality is typically too low for accurate FD analysis. Therefore, our study suggests using acceleration-based TD analysis for in vivo SWE to minimize underestimation of the true WS and, thus, to maximize the sensitivity of SWE to detect stiffness changes resulting from pathology.

Power Modulation Echocardiography to Detect and Quantify Myocardial Scar.

BACKGROUND: Myocardial scar correlates with clinical outcomes. Traditionally, late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) is used to detect and quantify scar. In this prospective study using LGE CMR as reference, the authors hypothesized that nonlinear ultrasound imaging, namely, power modulation, can detect and quantify myocardial scar in selected patients with previous myocardial infarction. In addition, given the different histopathology between ischemic and nonischemic scar, a further aim was to test the diagnostic performance of this echocardiographic technique in unselected consecutive individuals with ischemic and nonischemic LGE or no LGE on CMR. METHODS: Seventy-one patients with previous myocardial infarction underwent power modulation echocardiography following CMR imaging (group A). Subsequently, 101 consecutive patients with or without LGE on CMR, including individuals with nonischemic LGE, were scanned using power modulation echocardiography (group B). RESULTS: In group A, echocardiography detected myocardial scar in all 71 patients, with good scar volume agreement with CMR (bias = -1.9 cm3; limits of agreement [LOA], -8.0 to 4.2 cm3). On a per-segment basis, sensitivity was 82%, specificity 97%, and accuracy 92%. Sensitivity was higher in the inferior and posterior segments and lower in the anterior and lateral walls. In group B, on a per-subject basis, the sensitivity of echocardiography was 62% (91% for ischemic and 30% for nonischemic LGE), with specificity and accuracy of 89% and 72%, respectively. The bias for scar volume between modalities was 5.9 cm3, with LOA of 34.6 to 22.9 cm3 (bias = -1.9 cm3 [LOA, -11.4 to 7.6 cm3] for ischemic LGE, and bias = 18.9 cm3 [LOA, -67.4 to 29.7.6 cm3] for nonischemic LGE). CONCLUSIONS: Power modulation echocardiography can detect myocardial scar in both selected and unselected individuals with previous myocardial infarction and has good agreement for scar volume quantification with CMR. In an unselected cohort with nonischemic LGE, sensitivity is low.

Deep learning for dose assessment in radiotherapy by the super-localization of vaporized nanodroplets in high frame rate ultrasound imaging.

Objective.External beam radiotherapy is aimed to precisely deliver a high radiation dose to malignancies, while optimally sparing surrounding healthy tissues. With the advent of increasingly complex treatment plans, the delivery should preferably be verified by quality assurance methods. Recently, online ultrasound imaging of vaporized radiosensitive nanodroplets was proposed as a promising tool forin vivodosimetry in radiotherapy. Previously, the detection of sparse vaporization events was achieved by applying differential ultrasound (US) imaging followed by intensity thresholding using subjective parameter tuning, which is sensitive to image artifacts.Approach. A generalized deep learning solution (i.e. BubbleNet) is proposed to localize vaporized nanodroplets on differential US frames, while overcoming the aforementioned limitation. A 5-fold cross-validation was performed on a diversely composed 5747-frame training/validation dataset by manual segmentation. BubbleNet was then applied on a test dataset of 1536 differential US frames to evaluate dosimetric features. The intra-observer variability was determined by scoring the Dice similarity coefficient (DSC) on 150 frames segmented twice. Additionally, the BubbleNet generalization capability was tested on an external test dataset of 432 frames acquired by a phased array transducer at a much lower ultrasound frequency and reconstructed with unconventional pixel dimensions with respect to the training dataset.Main results.The median DSC in the 5-fold cross validation was equal to ∼0.88, which was in line with the intra-observer variability (=0.86). Next, BubbleNet was employed to detect vaporizations in differential US frames obtained during the irradiation of phantoms with a 154 MeV proton beam or a 6 MV photon beam. BubbleNet improved the bubble-count statistics by ∼30% compared to the earlier established intensity-weighted thresholding. The proton range was verified with a -0.8 mm accuracy.Significance.BubbleNet is a flexible tool to localize individual vaporized nanodroplets on experimentally acquired US images, which improves the sensitivity compared to former thresholding-weighted methods.

Cardiac troponin T and NT-proBNP for detecting myocardial ischemia in suspected chronic coronary syndrome.

BACKGROUND: Elevated N-terminal pro-B-type natriuretic peptides (NT-proBNP) and cardiac troponin T (cTnT) are associated with poor outcome in patients with chronic coronary syndrome (CCS). The performance of these biomarkers in diagnosing ischemia, and their association with myocardial hypoperfusion and hypokinesis is unclear. METHODS: Patients with suspected CCS (history of angina, estimated cardiovascular risk >15% or a positive stress test) were included in the prospective, multi-center DOPPLER-CIP study. Patients underwent Single Positron Emission Computed Tomography for assessment of ischemia and NT-proBNP and cTnT were measured in venous blood samples. RESULTS: We included 430 patients (25% female) aged 64 ± 8 years. Reversible hypoperfusion and hypokinesis were present in 139 (32%) and 89 (21%), respectively. Concentrations of NT-proBNP and cTnT correlated moderately (rho = 0.50, p < 0.001). NT-proBNP and cTnT concentrations (median [IQR]) were higher in patients with versus without reversible ischemia: 150 (73-294) versus 87 (44-192) ng/L and 10 (6-13) versus 7 (4-11) ng/L, respectively (p < 0.001 for both), and the associations persisted after adjusting for possible confounders. The C-statistics to discriminate ischemia ranged from 63%-73%, were comparable for cTnT and NT-proBNP, and higher for hypokinesis than hypoperfusion, and both were superior to exercise electrocardiography and stress echocardiography. Very low concentrations (≤5 ng/L cTnT and ≤ 60 ng/L NT-proBNP) ruled out reversible hypokinesis with negative predictive value >90%. CONCLUSION: cTnT and NT-proBNP are associated with irreversible and reversible ischemia in patients with suspected CCS, particularly hypokinesis. The diagnostic performance was comparable between the biomarkers, and very low concentrations may reliably rule out ischemia.

Ultrasound-assisted carbon ion dosimetry and range measurement using injectable polymer-shelled phase-change nanodroplets: in vitro study.

Methods allowing for in situ dosimetry and range verification are essential in radiotherapy to reduce the safety margins required to account for uncertainties introduced in the entire treatment workflow. This study suggests a non-invasive dosimetry concept for carbon ion radiotherapy based on phase-change ultrasound contrast agents. Injectable nanodroplets made of a metastable perfluorobutane (PFB) liquid core, stabilized with a crosslinked poly(vinylalcohol) shell, are vaporized at physiological temperature when exposed to carbon ion radiation (C-ions), converting them into echogenic microbubbles. Nanodroplets, embedded in tissue-mimicking phantoms, are exposed at 37 °C to a 312 MeV/u clinical C-ions beam at different doses between 0.1 and 4 Gy. The evaluation of the contrast enhancement from ultrasound imaging of the phantoms, pre- and post-irradiation, reveals a significant radiation-triggered nanodroplets vaporization occurring at the C-ions Bragg peak with sub-millimeter shift reproducibility and dose dependency. The specific response of the nanodroplets to C-ions is further confirmed by varying the phantom position, the beam range, and by performing spread-out Bragg peak irradiation. The nanodroplets' response to C-ions is influenced by their concentration and is dose rate independent. These early findings show the ground-breaking potential of polymer-shelled PFB nanodroplets to enable in vivo carbon ion dosimetry and range verification.

High-Frame-Rate Speckle Tracking for Echocardiographic Stress Testing.

Stress echocardiography helps to diagnose cardiac diseases that cannot easily be detected or do not even manifest at rest. In clinical practice, assessment of the stress test is usually performed visually and, therefore, in a qualitative and subjective way. Although speckle tracking echocardiography (STE) has been proposed for the quantification of function during stress, its time resolution is inadequate at high heart rates. Recently, high-frame-rate (HFR) imaging approaches have been proposed together with dedicated STE algorithms capable of handling small interframe displacements. The aim of this study was to determine if HFR STE is effective in assessing strain and strain rate parameters during echocardiographic stress testing. Specifically, stress echocardiography, at four different workload intensities, was performed in 25 healthy volunteers. At each stress level, HFR images from the apical four-chamber view were recorded using the ULA-OP 256 experimental scanner. Then, the myocardium was tracked with HFR STE, and strain and strain rate biomarkers were extracted to further analyze systolic and diastolic (early and late) peaks, as well as a short-lived isovolumic relaxation peak during stress testing. The global systolic strain response was monophasic, revealing a significant (p < 0.001) increase at low stress but then reaching a plateau. In contrast, all strain rate indices linearly increased (p < 0.001) with increasing stress level. These findings are in line with those reported using tissue Doppler imaging and, thus, indicate that HFR STE can be a useful tool in assessing cardiac function during stress echocardiography.

Acoustic Modulation Enables Proton Detection With Nanodroplets at Body Temperature.

Superheated nanodroplet (ND) vaporization by proton radiation was recently demonstrated, opening the door to ultrasound-based in vivo proton range verification. However, at body temperature and physiological pressures, perfluorobutane nanodroplets (PFB-NDs), which offer a good compromise between stability and radiation sensitivity, are not directly sensitive to primary protons. Instead, they are vaporized by infrequent secondary particles, which limits the precision for range verification. The radiation-induced vaporization threshold (i.e., sensitization threshold) can be reduced by lowering the pressure in the droplet such that ND vaporization by primary protons can occur. Here, we propose to use an acoustic field to modulate the pressure, intermittently lowering the proton sensitization threshold of PFB-NDs during the rarefactional phase of the ultrasound wave. Simultaneous proton irradiation and sonication with a 1.1 MHz focused transducer, using increasing peak negative pressures (PNPs), were applied on a dilution of PFB-NDs flowing in a tube, while vaporization was acoustically monitored with a linear array. Sensitization to primary protons was achieved at temperatures between [Formula: see text] and 40 °C using acoustic PNPs of relatively low amplitude (from 800 to 200 kPa, respectively), while sonication alone did not lead to ND vaporization at those PNPs. Sensitization was also measured at the clinically relevant body temperature (i.e., 37 °C) using a PNP of 400 kPa. These findings confirm that acoustic modulation lowers the sensitization threshold of superheated NDs, enabling a direct proton response at body temperature.

Spatially Variant Ultrasound Attenuation Mapping Using a Regularized Linear Least-Squares Approach.

Quantitative ultrasound methods aim to estimate the acoustic properties of the underlying medium, such as the attenuation and backscatter coefficients, and have applications in various areas including tissue characterization. In practice, tissue heterogeneity makes the coefficient estimation challenging. In this work, we propose a computationally efficient algorithm to map spatial variations of the attenuation coefficient. Our proposed approach adopts a fast, linear least-squares strategy to fit the signal model to data from pulse-echo measurements. As opposed to existing approaches, we directly estimate the attenuation map, that is, the local attenuation coefficient at each axial location by solving a joint estimation problem. In particular, we impose a physical model that couples all these local estimates and combine it with a smooth regularization to obtain a smooth map. Compared to the conventional spectral log difference method and the more recent ALGEBRA approach, we demonstrate that the attenuation estimates obtained by our method are more accurate and better correlate with the ground-truth attenuation profiles over a wide range of spatial and contrast resolutions.

MITEA: A dataset for machine learning segmentation of the left ventricle in 3D echocardiography using subject-specific labels from cardiac magnetic resonance imaging.

Segmentation of the left ventricle (LV) in echocardiography is an important task for the quantification of volume and mass in heart disease. Continuing advances in echocardiography have extended imaging capabilities into the 3D domain, subsequently overcoming the geometric assumptions associated with conventional 2D acquisitions. Nevertheless, the analysis of 3D echocardiography (3DE) poses several challenges associated with limited spatial resolution, poor contrast-to-noise ratio, complex noise characteristics, and image anisotropy. To develop automated methods for 3DE analysis, a sufficiently large, labeled dataset is typically required. However, ground truth segmentations have historically been difficult to obtain due to the high inter-observer variability associated with manual analysis. We address this lack of expert consensus by registering labels derived from higher-resolution subject-specific cardiac magnetic resonance (CMR) images, producing 536 annotated 3DE images from 143 human subjects (10 of which were excluded). This heterogeneous population consists of healthy controls and patients with cardiac disease, across a range of demographics. To demonstrate the utility of such a dataset, a state-of-the-art, self-configuring deep learning network for semantic segmentation was employed for automated 3DE analysis. Using the proposed dataset for training, the network produced measurement biases of -9 ± 16 ml, -1 ± 10 ml, -2 ± 5 %, and 5 ± 23 g, for end-diastolic volume, end-systolic volume, ejection fraction, and mass, respectively, outperforming an expert human observer in terms of accuracy as well as scan-rescan reproducibility. As part of the Cardiac Atlas Project, we present here a large, publicly available 3DE dataset with ground truth labels that leverage the higher resolution and contrast of CMR, to provide a new benchmark for automated 3DE analysis. Such an approach not only reduces the effect of observer-specific bias present in manual 3DE annotations, but also enables the development of analysis techniques which exhibit better agreement with CMR compared to conventional methods. This represents an important step for enabling more efficient and accurate diagnostic and prognostic information to be obtained from echocardiography.