Background and Purpose—: Carotid plaque rupture is a major cause of stroke. Key issue for risk stratification is early identification of rupture-prone plaques. A noninvasive technique, compound ultrasound strain imaging, was developed providing high-resolution radial deformation/strain images of atherosclerotic plaques. This study aims at in vivo validation of compound ultrasound strain imaging in patients by relating the measured strains to typical features of vulnerable plaques derived from histology after carotid endarterectomy. Materials and Methods—: Strains were measured in 34 severely stenotic (>70%) carotid arteries at the culprit lesion site within 48 hours before carotid endarterectomy. In all cases, the lumen-wall boundary was identifiable on B-mode ultrasound, and the imaged cross-section did not move out of the imaging plane from systole to diastole. After endarterectomy, the plaques were processed using a validated histology analysis technique. Results—: Locally elevated strain values were observed in regions containing predominantly components related to plaque vulnerability, whereas lower values were observed in fibrous, collagen-rich plaques. The median strain of the inner plaque layer (1 mm thickness) was significantly higher ( P <0.01) for (fibro)atheromatous (n=20, strain=0.27%) than that for fibrous plaques (n=14, strain=−0.75%). Also, a significantly larger area percentage of the inner layer revealed strains above 0.5% for (fibro)atheromatous (45.30%) compared with fibrous plaques (31.59%). (Fibro)atheromatous plaques were detected with a sensitivity, specificity, positive predictive value, and negative predictive value of 75%, 86%, 88%, and 71%, respectively. Strain did not significantly correlate with fibrous cap thickness, smooth muscle cell, or macrophage concentration. Conclusions—: Compound ultrasound strain imaging allows differentiating (fibro)atheromatous from fibrous carotid artery plaques.

Breast cancer is one of the most diagnosed types of cancer worldwide. Volumetric ultrasound breast imaging, combined with MRI can improve lesion detection rate, reduce examination time, and improve lesion diagnosis. However, to our knowledge, there are no 3D US breast imaging systems available that facilitate 3D US - MRI image fusion. In this paper, a novel Automated Cone-based Breast Ultrasound System (ACBUS) is introduced. The system facilitates volumetric ultrasound acquisition of the breast in a prone position without deforming it by the US transducer. Quality of ACBUS images for reconstructions at different voxel sizes (0.25 and 0.50 mm isotropic) was compared to quality of the Automated Breast Volumetric Scanner (ABVS) (Siemens Ultrasound, Issaquah, WA, USA) in terms of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and resolution using a custom made phantom. The ACBUS image data were registered to MRI image data utilizing surface matching and the registration accuracy was quantified using an internal marker. The technology was also evaluated in vivo. The phantom-based quantitative analysis demonstrated that ACBUS can deliver volumetric breast images with an image quality similar to the images delivered by a currently commercially available Siemens ABVS. We demonstrate on the phantom and in vivo that ACBUS enables adequate MRI-3D US fusion. To our conclusion, ACBUS might be a suitable candidate for a second-look breast US exam, patient follow-up, and US guided biopsy planning.

Plaque rupture in atherosclerotic carotid arteries is a main cause of ischemic stroke and it is correlated with high plaque stresses. Hence, analyzing stress patterns is essential for plaque specific rupture risk assessment. However, the critical information of the multicomponent material properties of atherosclerotic carotid arteries is still lacking greatly. This work aims to characterize component-wise material properties of atherosclerotic human carotid arteries under (almost) physiological loading conditions. An inverse finite element modeling (iFEM) framework was developed to characterize fibrous intima and vessel wall material properties of 13 cross sections from five carotids. The novel pipeline comprised ex-vivo inflation testing, pre-clinical high frequency ultrasound for deriving plaque deformations, pre-clinical high-magnetic field magnetic resonance imaging, finite element modeling, and a sample efficient machine learning based Bayesian Optimization. The nonlinear Yeoh constants for the fibrous intima and wall layers were successfully obtained. The optimization scheme of the iFEM reached the global minimum with a mean error of 3.8% in 133 iterations on average. The uniqueness of the results were confirmed with the inverted Gaussian Process (GP) model trained during the iFEM protocol. The developed iFEM approach combined with the inverted GP model successfully predicted component-wise material properties of intact atherosclerotic human carotids ex-vivo under physiological-like loading conditions. We developed a novel iFEM framework for the nonlinear, component-wise material characterization of atherosclerotic arteries and utilized it to obtain human atherosclerotic carotid material properties. The developed iFEM framework has great potential to be advanced for patient-specific in-vivo application.

In this study, the performances of one-dimensional and two-dimensional least-squares strain estimators (LSQSE) are compared. Furthermore, the effects of kernel size are examined using simulated raw frequency data of a widely-adapted hard lesion/soft tissue model. The performances of both methods are assessed in terms of root-mean-squared errors (RMSE), elastographic signal-to-noise ratio (SNR_c) and contrast-to-noise ratio (CNR_c). RMSE analysis revealed that the 2D LSQSE yields better results for phased array data, especially for larger insonification angles. Using a 2D LSQSE enabled the processing of unfiltered displacement data, in particular for the lateral/horizontal strain components. The SNR_c and CNR_c analysis showed an improvement in precision and almost no loss in contrast using 2D LSQSE. However, the RMSE images for different kernel sizes revealed that the optimal 2D kernel size depends on the region-of-interest and showed that the LSQ kernel size should be limited to avoid loss in resolution.