Automatic Segmentation of Abdominal Aortic Aneurysms from Time-Resolved 3D Ultrasound Images Using Deep Learning.

Abdominal aortic aneurysms (AAAs) are rupture-prone dilatations of the aorta. In current clinical practice, the maximal diameter of AAAs is monitored with 2D ultrasound to estimate their rupture risk. Recent studies have shown that 3-dimensional and mechanical AAA parameters might be better predictors for aneurysm growth and rupture than the diameter. These parameters can be obtained with time-resolved 3D ultrasound (3D+t US), which requires robust and automatic segmentation of AAAs from 3D+t US. This study proposes and validates a deep learning (DL) approach for automatic segmentation of AAAs. 500 AAA patients were included for follow-up 3D+t US imaging, resulting in 2495 3D+t US images. Segmentation masks for model training were obtained using a conventional automatic segmentation algorithm ('nonDL'). Four different DL models were trained and validated by (1) comparison to CT and (2) reader scoring. Performance of the nonDL and different DL segmentation strategies were evaluated by comparing Hausdorff distance, Dice scores, accuracy, sensitivity, and specificity with a sign test. All DL models had higher median Dice scores, accuracy, and sensitivity (all p < 0.003) compared to nonDL segmentation. The full image-resolution model without data augmentation showed the highest median Dice score and sensitivity (p < 0.001). Applying the DL model on an independent test group produced fewer poor segmentation scores of 1 to 2 on a five-point scale (8% for DL, 18% for nonDL). This demonstrates that a robust and automatic segmentation algorithm for segmenting abdominal aortic aneurysms from 3D+t US images was developed, showing improved performance compared to conventional segmentation.

In vivo bistatic dual-aperture ultrasound imaging and elastography of the abdominal aorta.

Introduction: In this paper we introduce in vivo multi-aperture ultrasound imaging and elastography of the abdominal aorta. Monitoring of the geometry and growth of abdominal aortic aneurysms (AAA) is paramount for risk stratification and intervention planning. However, such an assessment is limited by the lateral lumen-wall contrast and resolution of conventional ultrasound. Here, an in vivo dual-aperture bistatic imaging approach is shown to improve abdominal ultrasound and strain imaging quality significantly. By scanning the aorta from different directions, a larger part of the vessel circumference can be visualized. Methods: In this first-in-man volunteer study, the performance of multi-aperture ultrasound imaging and elastography of the abdominal aortic wall was assessed in 20 healthy volunteers. Dual-probe acquisition was performed in which two curved array transducers were aligned in the same imaging plane. The transducers alternately transmit and both probes receive simultaneously on each transmit event, which allows for the reconstruction of four ultrasound signals. Automatic probe localization was achieved by optimizing the coherence of the trans-probe data, using a gradient descent algorithm. Speckle-tracking was performed on the four individual bistatic signals, after which the respective axial displacements were compounded and strains were calculated. Results: Using bistatic multi-aperture ultrasound imaging, the image quality of the ultrasound images, i.e., the angular coverage of the wall, was improved which enables accurate estimation of local motion dynamics and strain in the abdominal aortic wall. The motion tracking error was reduced from 1.3 mm ± 0.63 mm to 0.16 mm ± 0.076 mm, which increased the circumferential elastographic signal-to-noise ratio (SNRe) by 12.3 dB ± 8.3 dB on average, revealing more accurate and homogeneous strain estimates compared to single-perspective ultrasound. Conclusion: Multi-aperture ultrasound imaging and elastography is feasible in vivo and can provide the clinician with vital information about the anatomical and mechanical state of AAAs in the future.

Increasing abdominal aortic aneurysm curvature visibility using 3D dual probe bistatic ultrasound imaging combined with probe translation.

High frame rate ultrasound (US) imaging techniques in 3D are promising tools for capturing abdominal aortic aneurysms (AAAs) over time, however, with the limited number of channel-to-element connections current footprints are small, which limits the field of view. Moreover, the maximal steering angle of the ultrasound beams in transmit and the maximal receptance angle in receive are insufficient for capturing the curvy shape of the AAA. Therefore, an approach is needed towards large arrays. In this study, high frame rate bistatic 3D US data (17 Hz) were acquired with two synchronized matrix arrays positioned at different locations (multi-aperture imaging) using a translation stage to simulate what a larger array with limited channel-to-element connections can potentially achieve. Acquisitions were performed along an AAA shaped phantom with different probe tilting angles (0 up to ± 30°). The performance of different multi-aperture configurations was quantified using the generalized contrast-to-noise ratio of the wall and lumen (gCNR). Furthermore, a parametric model of the multi-aperture system was used to estimate in which AAA wall regions the contrast is expected to be high. This was evaluated for AAAs with increasing diameters and curvature. With an eight-aperture 0° probe angle configuration a 69 % increase in field of view was measured in the longitudinal direction compared to the field of view of a single aperture configuration. When increasing the number of apertures from two to eight, the gCNR improved for the upper wall and lower wall by 35 % and 13 % (monostatic) and by 36 % and 13 % (bistatic). Contrast improvements up to 22 % (upper wall) and 12 % (lower wall) are achieved with tilted probe configurations compared to non-tilted configurations. Moreover, with bistatic imaging with tilted probe configurations gCNR improvements up to 4 % (upper wall) and 7 % (lower wall) are achieved compared to monostatic imaging. Furthermore, imaging with a larger inter-probe distance improved the gCNR for a ± 15° probe angle configuration. The gCNR has an expected pattern over time, where the contrast is lower when there is more wall motion (systole) and higher when motion is reduced (diastole). Furthermore, a higher frame rate (45 Hz) yields a lower gCNR, because fewer compound angles are used. The results of the parametric model suggest that a flat array is suitable for imaging AAA shapes with limited curvature, but that it is not suitable for imaging larger AAA shapes with more curvature. According to the model, tilted multi-aperture configurations combined with bistatic imaging can achieve a larger region with high contrast compared to non-tilted configurations. The findings of the model are in agreement with experimental findings. To conclude, this study demonstrates the vast improvements in field of view and AAA wall visibility that a large, sparsely populated 3D array can potentially achieve when imaging AAAs compared to single or dual aperture imaging. In the future, larger arrays, less thermal noise, more steering, and more channel-to-element connections combined with carefully chosen orientations of (sub-) apertures will likely advance 3D imaging of AAAs.

Deep learning assisted classification of spectral photoacoustic imaging of carotid plaques.

Spectral photoacoustic imaging (sPAI) is an emerging modality that allows real-time, non-invasive, and radiation-free assessment of tissue, benefiting from their optical contrast. sPAI is ideal for morphology assessment in arterial plaques, where plaque composition provides relevant information on plaque progression and its vulnerability. However, since sPAI is affected by spectral coloring, general spectroscopy unmixing techniques cannot provide reliable identification of such complicated sample composition. In this study, we employ a convolutional neural network (CNN) for the classification of plaque composition using sPAI. For this study, nine carotid endarterectomy plaques were imaged and were then annotated and validated using multiple histological staining. Our results show that a CNN can effectively differentiate constituent regions within plaques without requiring fluence or spectra correction, with the potential to eventually support vulnerability assessment in plaques.

Automatic Probe Localization in Multiaperture Ultrasound by Radon Domain Signal Matching.

In multiaperture ultrasound, several ultrasound probes with different insonification angles are combined to increase the field of view and angular coverage of image structures. A full reconstruction incorporating all possible combinations of transmitting and receiving probes has been shown to improve resolution, contrast, and angular coverage beyond what can be achieved by the registration of single images from different probes. A major challenge in multiaperture imaging is the correct determination of relative probe locations. A registration based on the content of images from different probes is challenging due to the decorrelation of image structures and speckle with increasing angle between the probes. We propose a probe localization method for plane-wave ultrasound that uses solely the receive dataset of a nontransmitting probe. The localization is performed by signal tracking in the Radon domain. To demonstrate that the method does not rely on common structures in the individual images, we show that a satisfying localization can be performed in pure speckle for angles, where the speckle patterns have completely decorrelated. The method shows potential for real-time probe localization in free-hand multiprobe ultrasound imaging or for flexible and wearable multiarray combination of multiple capacitive micromachined (CMUT)-based systems in the future.

Enabling strain imaging in realistic Eulerian ultrasound simulation methods.

Cardiovascular strain imaging is continually improving due to ongoing advances in ultrasound acquisition and data processing techniques. The phantoms used for validation of new methods are often burdensome to make and lack flexibility to vary mechanical and acoustic properties. Simulations of US imaging provide an alternative with the required flexibility and ground truth strain data. However, the current Lagrangian US strain imaging models cannot simulate heterogeneous speed of sound distributions and higher-order scattering, which limits the realism of the simulations. More realistic Eulerian modelling techniques exist but have so far not been used for strain imaging. In this research, a novel sampling scheme was developed based on a band-limited interpolation of the medium, which enables accurate strain simulation in Eulerian methods. The scheme was validated in k-Wave using various numerical phantoms and by a comparison with Field II. The method allows for simulations with a large range in strain values and was accurate with errors smaller than -60 dB. Furthermore, an excellent agreement with the Fourier theory of US scattering was found. The ability to perform simulations with heterogeneous speed of sound distributions was demonstrated using a pulsating artery model. The developed sampling scheme contributes to more realistic strain imaging simulations, in which the effect of heterogenous acoustic properties can be taken into account.

3D-Ultrasound Based Mechanical and Geometrical Analysis of Abdominal Aortic Aneurysms and Relationship to Growth.

The heterogeneity of progression of abdominal aortic aneurysms (AAAs) is not well understood. This study investigates which geometrical and mechanical factors, determined using time-resolved 3D ultrasound (3D + t US), correlate with increased growth of the aneurysm. The AAA diameter, volume, wall curvature, distensibility, and compliance in the maximal diameter region were determined automatically from 3D + t echograms of 167 patients. Due to limitations in the field-of-view and visibility of aortic pulsation, measurements of the volume, compliance of a 60 mm long region and the distensibility were possible for 78, 67, and 122 patients, respectively. Validation of the geometrical parameters with CT showed high similarity, with a median similarity index of 0.92 and root-mean-square error (RMSE) of diameters of 3.5 mm. Investigation of Spearman correlation between parameters showed that the elasticity of the aneurysms decreases slightly with diameter (p = 0.034) and decreases significantly with mean arterial pressure (p < 0.0001). The growth of a AAA is significantly related to its diameter, volume, compliance, and surface curvature (p < 0.002). Investigation of a linear growth model showed that compliance is the best predictor for upcoming AAA growth (RMSE 1.70 mm/year). To conclude, mechanical and geometrical parameters of the maximally dilated region of AAAs can automatically and accurately be determined from 3D + t echograms. With this, a prediction can be made about the upcoming AAA growth. This is a step towards more patient-specific characterization of AAAs, leading to better predictability of the progression of the disease and, eventually, improved clinical decision making about the treatment of AAAs.

Coherent Bistatic 3-D Ultrasound Imaging Using Two Sparse Matrix Arrays.

In the last decade, many advances have been made in high frame rate 3-D ultrasound imaging, including more flexible acquisition systems, transmit (TX) sequences, and transducer arrays. Compounding multiangle transmits of diverging waves has shown to be fast and effective for 2-D matrix arrays, where heterogeneity between transmits is key in optimizing the image quality. However, the anisotropy in contrast and resolution remains a drawback that cannot be overcome with a single transducer. In this study, a bistatic imaging aperture is demonstrated that consists of two synchronized matrix ( 32×32 ) arrays, allowing for fast interleaved transmits with a simultaneous receive (RX). First, for a single array, the aperture efficiency for high volume rate imaging was evaluated between sparse random arrays and fully multiplexed arrays. Second, the performance of the bistatic acquisition scheme was analyzed for various positions on a wire phantom and was showcased in a dynamic setup mimicking the human abdomen and aorta. Sparse array volume images were equal in resolution and lower in contrast compared to fully multiplexed arrays but can efficiently minimize decorrelation during motion for multiaperture imaging. The dual-array imaging aperture improved the spatial resolution in the direction of the second transducer, reducing the average volumetric speckle size with 72% and the axial-lateral eccentricity with 8%. In the aorta phantom, the angular coverage increased by a factor of 3 in the axial-lateral plane, raising the wall-lumen contrast with 16% compared to single-array images, despite accumulation of thermal noise in the lumen.

A Radon diffraction theorem for plane wave ultrasound imaging.

The rising demand on high frame rate ultrasound imaging applications necessitates the development of fast algorithms for plane wave image reconstruction. We introduce a new class of plane wave reconstructions that relies on a relation between receive data and image data in the Radon domain. This relation is derived for arbitrary dimensions and validated on multiple two-dimensional plane wave data sets. We further present a mathematical relation between conventional delay-and-sum and Fourier domain reconstruction methods and the method proposed. Our analysis shows that they all rely on the same physical model with slight variations in certain filtering steps and, therefore, the new Radon domain reconstruction yields similar results as other methods in terms of image quality. However, we show that our method offers a huge potential to improve computation time by reducing the number of applied projections and to improve image quality by introducing nonlinear operations in the Radon domain, e.g., for edge enhancement. As the Radon transform retains both angular and temporal information, the relation also provides new insights on the fundamentals of plane wave imaging that can be leveraged for optimizing acquisition schemes or for developing novel compounding strategies in the future.

Automatic Segmentation and Mechanical Characterisation of the Intraluminal Thrombus and Arterial Wall of Abdominal Aortic Aneurysms Using Time Resolved 3D Ultrasound Images.

OBJECTIVE: This study proposed a method for semi-automatic segmentation of abdominal aortic aneurysms (AAAs) and their intraluminal thrombus (ILT), based on time resolved 3D ultrasound (US), and validated results with computed tomography (CT). Mechanical properties of both wall and ILT were determined, and possible correlations with ILT size and blood pressure were investigated. METHODS: A semi-automatic segmentation algorithm was developed combining a star-Kalman approach with a 3D snake algorithm. The segmented geometries of both lumen and inner vessel wall were validated with both manual US based segmentations and CT based segmentations. Finally, the lumen and vessel wall distensibility and ILT compressibility were estimated and correlated with ILT size and blood pressure. RESULTS: For the vessel wall and lumen, the median Similarity Index (SI) was 92% (IQR 90, 94%) and 83% (IQR 75, 87%), respectively. The distensibility of the vessel wall could be determined in 37 of 40 cases and had a median value of 0.28 10-5 Pa-1 (IQR 0.18, 0.51 ×10-5). The median systolic to diastolic volume change of the ILT was determined successfully in 21 of 40 patients, and was -0.57% (IQR -1.1, 1.2%). The vessel and lumen distensibility showed a strong correlation with the systolic pressure (p < .010), rather than with the diastolic pressure. Lumen distensibility was strongly correlated with ILT thickness (p = .023). The performance of the semi-automatic segmentation algorithm was shown to be as good as the manual segmentations and highly dependent on the visibility of the ILT (limited contrast in seven patients and clutter in nine patients). CONCLUSION: This study has shown promising results for mechanical characterisation of the vessel, and ILT, including a correlation between distensibility, ILT size, and blood pressure. For future work, the inclusion rate needs to be increased by improving the image contrast with novel US techniques.

Blind spectral unmixing for characterization of plaque composition based on multispectral photoacoustic imaging.

To improve the assessment of carotid plaque vulnerability, a comprehensive characterization of their composition is paramount. Multispectral photoacoustic imaging (MSPAI) can provide plaque composition based on their absorption spectra. However, although various spectral unmixing methods have been developed to characterize different tissue constituents, plaque analysis remains a challenge since its composition is highly complex and diverse. In this study, we employed an adapted piecewise convex multiple-model endmember detection method to identify carotid plaque constituents. Additionally, we explore the selection of the imaging wavelengths in linear models by conditioning the coefficient matrix and its synergy with our unmixing approach. We verified our method using plaque mimicking phantoms and performed ex-vivo MSPAI on carotid endarterectomy samples in a spectral range from 500 to 1300 nm to identify the main spectral features of plaque materials for vulnerability assessment. After imaging, the samples were processed for histological analysis to validate the photoacoustic decomposition. Results show that our approach can perform spectral unmixing and classification of highly heterogeneous biological samples without requiring an extensive fluence correction, enabling the identification of relevant components to assess plaque vulnerability.

Contactless Cardiovascular Assessment by Imaging Photoplethysmography: A Comparison with Wearable Monitoring.

Despite the notable recent developments in the field of remote photoplethysmography (rPPG), extracting a reliable pulse rate variability (PRV) signal still remains a challenge. In this study, eight image-based photoplethysmography (iPPG) extraction methods (GRD, AGRD, PCA, ICA, LE, SPE, CHROM, and POS) were compared in terms of pulse rate (PR) and PRV features. The algorithms were made robust for motion and illumination artifacts by using ad hoc pre- and postprocessing steps. Then, they were systematically tested on the public dataset UBFC-RPPG, containing data from 42 subjects sitting in front of a webcam (30 fps) while playing a time-sensitive mathematical game. The performances of the algorithms were evaluated by statistically comparing iPPG-based and finger-PPG-based PR and PRV features in terms of Spearman's correlation coefficient, normalized root mean square error (NRMSE), and Bland-Altman analysis. The study revealed POS and CHROM techniques to be the most robust for PR estimation and the assessment of overall autonomic nervous system (ANS) dynamics by using PRV features in time and frequency domains. Furthermore, we demonstrated that a reliable characterization of the vagal tone is made possible by computing the Poincaré map of PRV series derived from the POS and CHROM methods. This study supports the use of iPPG systems as promising tools to obtain clinically useful and specific information about ANS dynamics.

An aberration correction approach for single and dual aperture ultrasound imaging of the abdomen.

Abdominal ultrasound image quality is hampered by phase aberration, that is mainly caused by the large speed-of-sound (SoS) differences between fat and muscle tissue in the abdominal wall. The mismatch between the assumed and actual SoS distribution introduces general blurring of the ultrasound images, and acoustic refraction can lead to geometric distortion of the imaged features. Large aperture imaging or dual-transducer imaging can improve abdominal imaging at deep locations by providing increased contrast and resolution. However, aberration effects for large aperture imaging can be even more severe, which limits its full potential. In this study, a model-based aberration correction method for arbitrary acquisition schemes is introduced for delay-and-sum (DAS) beamforming and its performance was analyzed for both single- and dual-transducer ultrasound imaging. The method employs aberration corrected wavefront arrival times, using manually assigned local SoS values. Two wavefront models were compared. The first model is based on a straight ray approximation, and the second model on the Eikonal equation, which is solved by a multi-stencils fast marching method. Their accuracy for abdominal imaging was evaluated in acoustic simulations and phantom experiments involving tissue-mimicking and porcine material with large SoS contrast (∼100 m/s). The lateral resolution was improved by up to 90% in simulations and up to 65% in experiments compared to standard DAS, in which the use of Eikonal beamforming generally outperformed straight ray beamforming. Moreover, geometric distortions were mitigated in multi-aperture imaging, leading to a reduction in position error of around 80%. A study on the sensitivity of the aberration correction to shape and SoS of aberrating layers was performed, showing that even with imperfect segmentations or SoS values, aberration correction still outperforms standard DAS.

The Impact of a Limited Field-of-View on Computed Hemodynamics in Abdominal Aortic Aneurysms: Evaluating the Feasibility of Completing Ultrasound Segmentations with Parametric Geometries.

To improve abdominal aortic aneurysm (AAA) rupture risk assessment, a large, longitudinal study on AAA hemodynamics and biomechanics is necessary, using personalized fluid-structure interaction (FSI) modeling. 3-dimensional, time-resolved ultrasound (3D+t US) is the preferred image modality to obtain the patient-specific AAA geometry for such a study, since it is safe, affordable and provides temporal information. However, the 3D+t US field-of-view (FOV) is limited and therefore often fails to capture the inlet and aorto-iliac bifurcation geometry. In this study, a framework was developed to add parametric inlet and bifurcation geometries to the abdominal aortic aneurysm geometry by employing dataset statistics and parameters of the AAA geometry. The impact of replacing the patient-specific inlet and bifurcation geometries, acquired using computed tomography (CT) scans, by parametric geometries was evaluated by examining the differences in hemodynamics (systolic and time-averaged wall shear stress and oscillatory shear index) in the aneurysm region. The results show that the inlet geometry has a larger effect on the AAA hemodynamics (median differences of 7.5 to 18.8%) than the bifurcation geometry (median differences all below 1%). Therefore, it is not feasible to replace the patient-specific inlet geometry by a generic one. Future studies should investigate the possibilities of extending the proximal FOV of 3D+t US. However, this study did show the feasibility of adding a parametric bifurcation geometry to the aneurysm geometry. After extending the proximal FOV, the obtained framework can be used to extract AAA geometries from 3D+t US for FSI simulations, despite the absence of the bifurcation geometry.

Spatiotemporal Registration of 3-D Multi-perspective Ultrasound Images of Abdominal Aortic Aneurysms.

Methods for patient-specific abdominal aortic aneurysm (AAA) progression monitoring and rupture risk assessment are widely investigated. Three-dimensional ultrasound can visualize the AAA's complex geometry and displacement fields. However, ultrasound has a limited field of view and low frame rate (i.e., 3-8 Hz). This article describes an approach to enhance the temporal resolution and the field of view. First, the frame rate was increased for each data set by sequencing multiple blood pulse cycles into one cycle. The sequencing method uses the original frame rate and the estimated pulse wave rate obtained from AAA distension curves. Second, the temporal registration was applied to multi-perspective acquisitions of the same AAA. Third, the field of view was increased through spatial registration and fusion using an image feature-based phase-only correlation method and a wavelet transform, respectively. Temporal sequencing was fully correct in aortic phantoms and was successful in 51 of 62 AAA patients, yielding a factor 5 frame rate increase. Spatial registration of proximal and distal ultrasound acquisitions was successful in 32 of 37 different AAA patients, based on the comparison between the fused ultrasound and computed tomography segmentation (95th percentile Haussdorf distances and similarity indices of 4.2 ± 1.7 mm and 0.92 ± 0.02 mm, respectively). Furthermore, the field of view was enlarged by 9%-49%.

Multiperspective Photoacoustic Imaging Using Spatially Diverse CMUTs.

Photoacoustic imaging (PAI) is a promising technique to assess different constituents in tissue. In PAI, the propagating waves are low-amplitude, isotropic, and broadband. A common approach in PAI is the use of a single linear or curved piezoelectric transducer array to perform both PA and ultrasound imaging. These systems provide freedom, agility, and versatility for performing imaging, but have limited field of view (FOV) and directivity that degrade the final image quality. Capacitive micromachined ultrasonic transducers (CMUTs) have a great potential to be used for PAI since they provide larger bandwidth and better cost efficiency. In this study, to improve the FOV, resolution, and contrast, we propose a multiperspective PAI (MP-PAI) approach using multiple CMUTs on a flexible array with shared channels. The designed array was used to perform MP-PAI in an in vitro experiment using a plaque mimicking phantom where the images were compounded both incoherently and coherently. The MP-PAI approach showed a significant improvement in overall image quality. Using only three CMUTs led to about 20% increase in generalized-contrast-to-noise ratio (gCNR), 2-dB improvement in peak signal-to-noise ratio (PSNR), and double the structural coverage in comparison to a single CMUT setup. In numerical studies, the MP-PAI was thoroughly evaluated for both the coherent and incoherent compounding methods. The assessments showed that the image quality further improved for increased number of transducers and angular coverage. For 15 transducers, the improvement for resolution and contrast could be up to three times the amount in a single-perspective image. Nonetheless, the most prominent improvement of MP-PAI was its ability to resolve the structural information of the phantoms.

Characterization of intra-tissue strain fields in articular cartilage explants during post-loading recovery using high frequency ultrasound.

This study aims to demonstrate the potential of ultrasound elastography as a research tool for non-destructive imaging of intra-tissue strain fields and tissue quality assessment in cartilage explants. Osteochondral plugs from bovine patellae were loaded up to 10, 40, or 70 N using a hemi-spherical indenter. The load was kept constant for 15 min, after which samples were unloaded and ultrasound imaging of strain recovery over time was performed in the indented area for 1 h. Tissue strains were determined using speckle tracking and accumulated to LaGrangian strains in the indentation direction. For all samples, strain maps showed a heterogeneous strain field, with the highest values in the superficial cartilage under the indenter tip at the bottom of the indent and decreasing values in the deeper cartilage. Strains were higher at higher load levels and tissue recovery over time was faster after indentation at 10 N than at 40 N and 70 N. At lower compression levels most displacement occurred near the surface with little deformation in the deep layers, while at higher levels strains increased more evenly in all cartilage zones. Ultrasound elastography is a promising method for high resolution imaging of intra-tissue strain fields and evaluation of cartilage quality in tissue explants in a laboratory setting. In the future, it may become a clinical diagnostic tool used to identify the extent of cartilage damage around visible defects.

Modeling toolchain for realistic simulation of photoacoustic data acquisition.

SIGNIFICANCE: Physics-based simulations of photoacoustic (PA) signals are used to validate new methods, to characterize PA setups and to generate training datasets for machine learning. However, a thoroughly validated PA simulation toolchain that can simulate realistic images is still lacking. AIM: A quantitative toolchain was developed to model PA image acquisition in complex tissues, by simulating both the optical fluence and the acoustic wave propagation. APPROACH: Sampling techniques were developed to decrease artifacts in acoustic simulations. The performance of the simulations was analyzed by measuring the point spread function (PSF) and using a rotatable three-channel phantom, filled with cholesterol, a human carotid plaque sample, and porcine blood. Ex vivo human plaque samples were simulated to validate the methods in more complex tissues. RESULTS: The sampling techniques could enhance the quality of the simulated PA images effectively. The resolution and intensity of the PSF in the turbid medium matched the experimental data well. Overall, the appearance, signal-to-noise ratio and speckle of the images could be simulated accurately. CONCLUSIONS: A PA toolchain was developed and validated, and the results indicate a great potential of PA simulations in more complex and heterogeneous media.

Unsupervised domain adaptation method for segmenting cross-sectional CCA images.

BACKGROUND AND OBJECTIVES: Automatic vessel segmentation in ultrasound is challenging due to the quality of the ultrasound images, which is affected by attenuation, high level of speckle noise and acoustic shadowing. Recently, deep convolutional neural networks are increasing in popularity due to their great performance on image segmentation problems, including vessel segmentation. Traditionally, large labeled datasets are required to train a network that achieves high performance, and is able to generalize well to different orientations, transducers and ultrasound scanners. However, these large datasets are rare, given that it is challenging and time-consuming to acquire and manually annotate in-vivo data. METHODS: In this work, we present a model-based, unsupervised domain adaptation method that consists of two stages. In the first stage, the network is trained on simulated ultrasound images, which have an accurate ground truth. In the second stage, the network continues training on in-vivo data in an unsupervised way, therefore not requiring the data to be labelled. Rather than using an adversarial neural network, prior knowledge on the elliptical shape of the segmentation mask is used to detect unexpected outputs. RESULTS: The segmentation performance was quantified using manually segmented images as ground truth. Due to the proposed domain adaptation method, the median Dice similarity coefficient increased from 0 to 0.951, outperforming a domain adversarial neural network (median Dice 0.922) and a state-of-the-art Star-Kalman algorithm that was specifically designed for this dataset (median Dice 0.942). CONCLUSIONS: The results show that it is feasible to first train a neural network on simulated data, and then apply model-based domain adaptation to further improve segmentation performance by training on unlabeled in-vivo data. This overcomes the limitation of conventional deep learning approaches to require large amounts of manually labeled in-vivo data. Since the proposed domain adaptation method only requires prior knowledge on the shape of the segmentation mask, performance can be explored in various domains and applications in future research.

Ultrasound-Based Estimation of Fibre-Directional Strain: A Simulation Study.

Left ventricular (LV) strains are typically represented with respect to the imaging axes. Contraction within the myocardium occurs along myofibres, which vary in orientation. Therefore, a mismatch exists between the direction in which strain is calculated and that in which contraction occurs. In this study, ultrasound-based fibre orientation and 3-D strain estimation were combined to calculate the fibre-directional strain. Three-dimensional ultrasound volumes were created by simulating simple geometrical phantoms and a phantom based on a finite-element (FE) model of LV mechanics. Fibre-like structures were embedded within tissue-mimicking scatterers. Strains were applied to the numerical phantom, whereas the FE phantom was deformed based on the LV model. Fibre orientation was accurately estimated for both phantoms. There was poor agreement in axial and elevational strains (root mean square error = 29.9% and 12.3%), but good agreement in lateral and fibre-directional strains (root mean square error = 6.4% and 5.9% respectively), which aligned in the midwall. Simplifications to reduce computational complexity caused poor axial and elevational strain estimation. However, calculation of fibre-directional strain from single-modality ultrasound volumes was successful. Further studies, in ex vivo setups because of the fundamental limitations of currently available transducers, are needed to verify real-world performance of the method.