Standard B-mode Ultrasound Imaging Research Articles

Joint segmentation and classification of breast masses based on ultrasound radio-frequency data and convolutional neural networks

In this paper, we propose a novel deep learning method for joint classification and segmentation of breast masses based on radio-frequency (RF) ultrasound (US) data. In comparison to commonly used classification and segmentation techniques, utilizing B-mode US images, we train the network with RF data (data before envelope detection and dynamic compression), which are considered to include more information on tissue’s physical properties than standard B-mode US images. Our multi-task network, based on the Y-Net architecture, can effectively process large matrices of RF data by mixing 1D and 2D convolutional filters. We use data collected from 273 breast masses to compare the performance of networks trained with RF data and US images. The multi-task model developed based on the RF data achieved good classification performance, with area under the receiver operating characteristic curve (AUC) of 0.90. The network based on the US images achieved AUC of 0.87. In the case of the segmentation, we obtained mean Dice scores of 0.64 and 0.60 for the approaches utilizing US images and RF data, respectively. Moreover, the interpretability of the networks was studied using class activation mapping technique and by filter weights visualizations.

Estimation of absolute states of human skeletal muscle via standard B-mode ultrasound imaging and deep convolutional neural networks.

The objective is to test automated in vivo estimation of active and passive skeletal muscle states using ultrasonic imaging. Current technology (electromyography, dynamometry, shear wave imaging) provides no general, non-invasive method for online estimation of skeletal muscle states. Ultrasound (US) allows non-invasive imaging of muscle, yet current computational approaches have never achieved simultaneous extraction or generalization of independently varying active and passive states. We use deep learning to investigate the generalizable content of two-dimensional (2D) US muscle images. US data synchronized with electromyography of the calf muscles, with measures of joint moment/angle, were recorded from 32 healthy participants (seven female; ages: 27.5, 19–65). We extracted a region of interest of medial gastrocnemius and soleus using our prior developed accurate segmentation algorithm. From the segmented images, a deep convolutional neural network was trained to predict three absolute, drift-free components of the neurobiomechanical state (activity, joint angle, joint moment) during experimentally designed, simultaneous independent variation of passive (joint angle) and active (electromyography) inputs. For all 32 held-out participants (16-fold cross-validation) the ankle joint angle, electromyography and joint moment were estimated to accuracy 55 ± 8%, 57 ± 11% and 46 ± 9%, respectively. With 2D US imaging, deep neural networks can encode, in generalizable form, the activity–length–tension state relationship of these muscles. Observation-only, low-power 2D US imaging can provide a new category of technology for non-invasive estimation of neural output, length and tension in skeletal muscle. This proof of principle has value for personalized muscle assessment in pain, injury, neurological conditions, neuropathies, myopathies and ageing.

Acoustic wave sparsely activated localization microscopy (AWSALM): Super-resolution ultrasound imaging using acoustic activation and deactivation of nanodroplets

Photo-activated localization microscopy (PALM) has revolutionized the field of fluorescence microscopy by breaking the diffraction limit in spatial resolution. In this study, “acoustic wave sparsely activated localization microscopy (AWSALM),” an acoustic counterpart of PALM, is developed to super-resolve structures which cannot be resolved by conventional B-mode imaging. AWSALM utilizes acoustic waves to sparsely and stochastically activate decafluorobutane nanodroplets by acoustic vaporization and to simultaneously deactivate the existing vaporized nanodroplets via acoustic destruction. In this method, activation, imaging, and deactivation are all performed using acoustic waves. Experimental results show that sub-wavelength micro-structures not resolvable by standard B-mode ultrasound images can be separated by AWSALM. This technique is flow independent and does not require a low concentration of contrast agents, as is required by current ultrasound super resolution techniques. Acoustic activation and deactivation can be controlled by adjusting the acoustic pressure, which remains well within the FDA approved safety range. In conclusion, this study shows the promise of a flow and contrast agent concentration independent super-resolution ultrasound technique which has potential to be faster and go beyond vascular imaging.

Recent Developments in Multiparametric Prostate MR Imaging

Multiparametric magnetic resonance imaging (mpMRI) of the prostate has become a mature and accepted technology in the evaluation of the patient with suspected prostate cancer (PCa). This review focuses on key recent developments in the area of mpMRI, specifically: reporting systems used by radiologists when interpreting mpMRI, advances in technical aspects of diffusion-weighted imaging (DWI), and innovations in MRI-ultrasound (MRI-US) fusion-guided prostate biopsies. These topics were selected given that reporting systems are critical when communicating findings on mpMRI of the prostate to referring clinicians, advanced techniques in DWI have improved the detection and characterization of PCa, and MRI-US fusion biopsies now enable more accurate targeting of suspicious lesions at biopsy, including lesions that are difficult to visualize using standard B-mode ultrasound imaging. Radiologists interpreting prostate MRI today are likely to incorporate aspects of all three of these topics into their current practice.

Prospective phase I/II trial of dose-painting prostate brachytherapy guided by cancer-specific ultrasound-based tissue-type imaging (TTI).

246 Background: Standard treatments treat the entire prostate gland uniformly, while focal treatments target only the dominant tumor. We propose an alternative approach wherein tumor areas are treated with an escalated dose of radiation while the remainder of the prostate with dose de-escalation. Tissue-type images (TTIs), which identify intraprostatic tumors via ultrasound spectral analysis, were used to identify tumors. Methods: 14 low risk patients were enrolled. TTIs and standard B-mode ultrasound images were generated and fused intraoperatively. Based on TTIs, plans were developed to deliver 200% prescription dose to the tumors and 100% to the prostate. Doses above 100% to non-tumor regions were minimized and standard normal tissue constraints were respected. Results: 12 patients were successfully treated. Technical problems occurred in the first patient and another received a standard implant due to discordance of TTIs with MRI and biopsy findings. One patient passed away soon after the procedure of unrelated cause. A total of 28 tumors were identified in these 12 patients (median 2 tumors/patient, range 1-4). Mean tumor volume was 0.72cc (range 0.03cc -4.07cc). Tumor dose was significantly higher in TTI-based plans, with mean tumor V200%: TTI vs. standard 97% vs. 48%, p<0.05. Modest dose reduction of the whole prostate was also achieved. Median follow-up was 25.7 months (range 21.3-48 months). A low incidence of grade 2 toxicities and no grade 3-4 toxicities were seen (Table). Two patients experienced a biochemical failure per the nadir+2 definition, both failures are attributable to PSA bounce (PSA down to 1.0 and 0.59 at 42 and 30 mos., respectively). Post-treatment biopsies have been performed on 5/11 patients (4 refused), with 2 patients with pending biopsies. 4/5 available biopsies are negative, with 1 patient having a small focus of residual cancer. Conclusions: Dose-painting prostate brachytherapy is technically feasible. This tumor-directed approach, more rational than uniform whole gland treatments, offers the promise of excellent cancer control, low toxicity and low risk of re-treatment. Clinical trial information: NCT01227642. [Table: see text]

Integrated ultrasound and magnetic resonance imaging for simultaneous temperature and cavitation monitoring during focused ultrasound therapies

Ultrasound can be used to noninvasively produce different bioeffects via viscous heating, acoustic cavitation, or their combination, and these effects can be exploited to develop a wide range of therapies for cancer and other disorders. In order to accurately localize and control these different effects, imaging methods are desired that can map both temperature changes and cavitation activity. To address these needs, the authors integrated an ultrasound imaging array into an MRI-guided focused ultrasound (MRgFUS) system to simultaneously visualize thermal and mechanical effects via passive acoustic mapping (PAM) and MR temperature imaging (MRTI), respectively. The system was tested with an MRgFUS system developed for transcranial sonication for brain tumor ablation in experiments with a tissue mimicking phantom and a phantom-filled ex vivo macaque skull. In experiments on cavitation-enhanced heating, 10 s continuous wave sonications were applied at increasing power levels (30-110 W) until broadband acoustic emissions (a signature for inertial cavitation) were evident. The presence or lack of signal in the PAM, as well as its magnitude and location, were compared to the focal heating in the MRTI. Additional experiments compared PAM with standard B-mode ultrasound imaging and tested the feasibility of the system to map cavitation activity produced during low-power (5 W) burst sonications in a channel filled with a microbubble ultrasound contrast agent. When inertial cavitation was evident, localized activity was present in PAM and a marked increase in heating was observed in MRTI. The location of the cavitation activity and heating agreed on average after registration of the two imaging modalities; the distance between the maximum cavitation activity and focal heating was -3.4 ± 2.1 mm and -0.1 ± 3.3 mm in the axial and transverse ultrasound array directions, respectively. Distortions and other MRI issues introduced small uncertainties in the PAM∕MRTI registration. Although there was substantial variation, a nonlinear relationship between the average intensity of the cavitation maps, which was relatively constant during sonication, and the peak temperature rise was evident. A fit to the data to an exponential had a correlation coefficient (R(2)) of 0.62. The system was also found to be capable of visualizing cavitation activity with B-mode imaging and of passively mapping cavitation activity transcranially during cavitation-enhanced heating and during low-power sonication with an ultrasound contrast agent. The authors have demonstrated the feasibility of integrating an ultrasound imaging array into an MRgFUS system to simultaneously map localized cavitation activity and temperature. The authors anticipate that this integrated approach can be utilized to develop controllers for cavitation-enhanced ablation and facilitate the optimization and development of this and other ultrasound therapies. The integrated system may also provide a useful tool to study the bioeffects of acoustic cavitation.

Impairment in aortic elastic properties and mechanics of ascending aorta in patients with bicuspid aortic valve

Background: Bicuspid aortic valve (BAV), which is one of the most common congenital heart defects, is a genetic disorder of the aortic valve and ascending aorta. Aortic dilation is common in subjects with BAV. In the present study we aimed to investigate the elastic and mechanical properties of the aorta by using a novel strain method, velocity vector imaging (VVI), in patients with BAV. Methods: We studied 22 patients with BAV (age 43.31±4.5, 57% female) and 20 age and sex-matched healthy subjects with tricuspid aortic valve. All the patients and healthy controls were subjected to assessment of aortic strain, stiffness and distensibility measurements of M-mode images of the aorta. VVI measurements were obtained from off-line analysis of standard B-mode ultrasound images of the ascending aorta. Results: Aortic strain (5.2±1.5% to 13.91±4.77%, respectively, p=0.0001), and distensibility 0.19±0.06% to 1.6±0.77%, respectively, p=0.0001) were significantly impaired in patients with BAV, compared to the control group. Aortic stiffness was markedly increased in patients with BAV (4.45±1.73 to 2.6±1.3, respectively, p=0.001).Regarding VVI-based strain measurements, peak longitudinal strain (S), strain rate (SR) and total longitudinal displacement (TLD) values were significantly impaired in patients with BAV, compared to the controls. (S: 11.49±1.2% to 14.6±2.58%, p=0.03; SR: 0.57±0.03 1/s to 1.66±0.40 1/s, p=0.0001; TLD: 3.98±1.78 mm to 10.42±2.38 mm, p= 0.0001). Conclusions: BAV is associated with reduced elasticity of the aorta. Longitudinal wall motion of ascending aorta is also impaired in patients with BAV. VVI is a novel strain method which provides detailed data in assessing the longitudinal deformation of the ascending aorta, in patients with BAV.

SAT0399 Impairment in aortic elastic properties and mechanics of carotid artery system in patients with takayasu’ s arteritis

BackgroundTakayasu’s arteritis (TA) is a chronic inflammatory vasculitis of the aorta and its major branches. Increased arterial stiffness is known to be a predictor of cardiovascular mortality. Impairment of the...

Tendon Strain Measurements With Dynamic Ultrasound Images: Evaluation of Digital Image Correlation

Strain is an essential metric in tissue mechanics. Strains and strain distributions during functional loads can help identify damaged and pathologic regions as well as quantify functional compromise. Noninvasive strain measurement in vivo is difficult to perform. The goal of this in vitro study is to determine the efficacy of digital image correlation (DIC) methods to measure strain in B-mode ultrasound images. The Achilles tendons of eight male Wistar rats were removed and mechanically cycled between 0 and 1% strain. Three cine video images were captured for each specimen: (1) optical video for manual tracking of optical markers; (2) optical video for DIC tracking of optical surface markers; and (3) ultrasound video for DIC tracking of image texture within the tissue. All three imaging modalities were similarly able to measure tendon strain during cyclic testing. Manual/ImageJ-based strain values linearly correlated with DIC (optical marker)-based strain values for all eight tendons with a slope of 0.970. DIC (optical marker)-based strain values linearly correlated with DIC (ultrasound texture)-based strain values for all eight tendons with a slope of 1.003. Strain measurement using DIC was as accurate as manual image tracking methods, and DIC tracking was equally accurate when tracking ultrasound texture as when tracking optical markers. This study supports the use of DIC to calculate strains directly from the texture present in standard B-mode ultrasound images and supports the use of DIC for in vivo strain measurement using ultrasound images without additional markers, either artificially placed (for optical tracking) or anatomically in view (i.e., bony landmarks and/or muscle-tendon junctions).

Longitudinal wall motion of the common carotid artery can be assessed by velocity vector imaging

Velocity vector imaging (VVI) is novel ultrasound image analysis software, enabling simultaneous evaluation of longitudinal and radial tissue motion. This study aimed to investigate the possible usefulness of VVI in evaluating the longitudinal vessel wall movement of the common carotid artery (CCA). Sixteen healthy volunteers and 16 patients with established coronary artery disease (CAD) were included in the study. CCA was scanned and standard B-mode ultrasound images were analysed off-line with VVI. In healthy volunteers, total longitudinal displacements (tLoD) of the right and left CCA were similar, as were the movements of the near- and far wall of the right CCA. The CAD group showed significantly lower tLoD compared to the healthy volunteers (0·543 ± 0·394 versus 0·112 ± 0·074, P<0·0001). VVI is a highly feasible technique in assessing longitudinal CCA wall motion, which may be of potential relevance as a novel vascular biomarker.

A new time-gain correction method for standard B-mode ultrasound imaging

A method used to provide gain variation with received echo range to compensate for tissue attenuation in standard B-mode ultrasound imaging systems achieves signal multiplication by first logarithmically compressing the complete dynamic range of the received echo. The time-gain correction method described here is a modification of this method. It performs the time-gain correction operation digitally between the digital scan converter and the display. The overall system transfer function can be specified by means of a software interface without changing the characteristics of the analog front end. This interface can be automated to almost any degree, and it can include preset settings for typical imaging requirements. This method also allows the gain function to be changed even when new images are not being generated.