3D Ultrasound Imaging Research Articles

A Deep Reinforcement Learning Based Region-Specific Beamformer for Sparse Arrays 3D Ultrasound Imaging.

Sparse arrays offer several advantages over other element reduction techniques for 3D ultrasound imaging. However, the large inter-element spacing in these arrays results in high sidelobe-related artifacts, which significantly degrade image quality and limit their application in 3D ultrasound imaging. Adaptive beamformers have been proposed to mitigate sidelobe-related artifacts, but they often degrade speckle texture quality, resulting in unnaturally dark images. To overcome these limitations, we propose RSB-Net, a region-specific beamformer based on deep reinforcement learning. RSB-Net adaptively selects the most suitable beamformer for each pixel of the image, applying adaptive beamforming in regions dominated by sidelobe artifacts and delay-and-sum beamforming in regions where speckle texture should be preserved. The effectiveness of RSB-Net was validated on both simulated and experimental synthetic transmit aperture RF datasets with a new designed sparse array prototype. On simulated data, RSB-Net achieved significant gains, with improvements of 52.81 dB in contrast ratio and 0.65 in generalized contrast-to-noise ratio compared to DAS beamforming. In experimental tissue-mimicking phantom data, RSB-Net demonstrated similar performance, achieving gains of 51.01 dB and 0.64 respectively. These results highlight the potential of RSB-Net as a robust and effective solution for high-quality B-mode 3D ultrasound imaging using 2D sparse arrays, advancing the standardization of 3D ultrasound in clinical settings by enhancing anatomical visualization, reducing operator dependency, and improving measurement accuracy for lesions and calcifications.

Study on Ultrasound-Assisted Diagnosis of CHB Complicated with NAFLD Hepatic Fibrosis Based on Deep Learning.

This study aimed to develop and evaluate an automated classification model for diagnosing chronic hepatitis B (CHB) complicated with non-alcoholic fatty liver disease (NAFLD) using deep learning techniques applied to two-dimensional liver imaging. We retrospectively analyzed data from 2803 patients diagnosed with CHB and NAFLD via two-dimensional ultrasound and FibroScan at our hospital between June 2019 and December 2022. These patients contributed a total of 20,540 two-dimensional liver images. An additional 150 patients with CHB complicated with NAFLD, who had liver biopsy results, were selected (with a total of 922 liver 2D ultrasound images) for validation of the deep learning model. The diagnostic performance was assessed using sensitivity, specificity, and accuracy metrics for independent diagnoses made by clinicians or with AI assistance. Diagnostic performance was assessed under blinded conditions by physicians of varying expertise, primary, intermediate, and senior clinicians, either independently or assisted by the AI model. We then conducted a statistical analysis of the results, focusing on sensitivity, specificity, and accuracy metrics. The sensitivity, specificity, and accuracy for independent diagnoses made by primary, intermediate, and senior physicians were 24.3%, 29.1%, and 33.8% for sensitivity; 100% for specificity; and 25.3%, 30.0%, and 34.7% for accuracy, respectively. The AI model alone achieved 68.9% sensitivity, 100% specificity, and 69.3% accuracy. When diagnoses were AI-assisted, the sensitivity increased significantly to 73.7% for primary physicians, 73.7% for intermediate physicians, and 75.7% for senior physicians, while specificity remained at 100%, and accuracy was 74.0%, 74.0%, and 76.0%, respectively. The area under the receiver operating characteristic curve (AUC) for AI-independent diagnosis (0.845) was significantly higher than that for primary, intermediate, and senior physicians (0.622, 0.645, and 0.669, respectively; ). Similarly, AI-assisted diagnosis AUC values (0.868, 0.868, and 0.878, respectively) surpassed those for independent diagnosis by physicians at each level of expertise ( ). In conclusion, the deep learning model based on two-dimensional liver imaging demonstrated feasibility for assisting in the identification of liver fibrosis in patients with CHB and NAFLD. Our results suggest that integrating AI models into the diagnostic process can significantly improve the accuracy of diagnosing liver conditions in this patient population.

Anomalous drainage of the ductus venosus into the coronary sinus: prenatal ultrasound diagnosis utilizing two-dimensional and three-dimensional imaging techniques and differential diagnosis

BackgroundAnomalous drainage of the ductus venosus (DV) into the coronary sinus (CS) is a rare fetal vascular anomaly that poses challenges for prenatal diagnosis. This study aimed to investigate the role of two-dimensional (2D) and three-dimensional (3D) ultrasound imaging, specifically spatiotemporal image correlation (STIC) technology, in improving the prenatal diagnosis of this anomaly.MethodsWe retrospectively reviewed eight cases of fetuses diagnosed with anomalous DV drainage into the CS at Gansu Provincial Maternal and Child Health Hospital between September 2019 and September 2024. The gestational age of the fetuses ranged from 24 to 30 weeks. Ultrasound examinations, including 2D and Doppler imaging, along with HDlive Flow combined with STIC technology, were used for diagnosis. Differential diagnoses were made based on imaging findings. Descriptive statistics were employed to summarize the results.ResultsEight fetuses with anomalous DV drainage into the CS were identified. Of these, five cases were isolated anomalies, while three had associated malformations, such as aberrant right subclavian artery and right aortic arch. Dilated CS was observed in all cases, with an average inner diameter of 5.7 mm. STIC imaging successfully visualized the abnormal course of the DV, enhancing diagnostic confidence. Postnatal follow-up indicated favorable outcomes for most neonates (except for Case NO.3 and Case NO.7), although persistent CS dilation was observed in these cases, without significant hemodynamic compromise or clinical symptoms.ConclusionsPrenatal diagnosis of anomalous DV drainage into the CS can be effectively achieved using 2D and 3D ultrasound, with STIC technology providing added diagnostic clarity. Early and accurate detection is crucial for ensuring appropriate clinical management and favorable outcomes. Ongoing surveillance of CS dilation in the postnatal period is recommended.

Fetal Cerebellum Landmark Detection Based on 3D MRI: Method and Benchmark.

Fetal cerebellum landmark detection is crucial for assessing fetal brain development. Although deep learning has become the standard for automatic landmark detection, most previous methods have focused on using 2D ultrasound or thick Magnetic Resonance Imaging (MRI) . To improve accuracy, landmarks should be located on thin 3D MRIs. However, abnormal development, high noise, and fuzzy boundaries in 3D fetal brain images make traditional methods less effective for cerebellum landmark detection. To address this, we introduce the Anatomical Pseudo-label Guided Attention (APGA) network alongside a 3D MRI-based benchmark for fetal cerebellum landmark detection. During training, we use a shared encoder to extract image features and two decoders for landmark regression and anatomical pseudo-label segmentation. We design a Feature Decoupling Transformer (FDT) and embed it into the encoder to better calibrate the features for the two tasks. We only need the encoder, the FDT, and the landmark decoder during the inference phase. Extensive experiments on our proposed benchmark and out-of-domain test set have shown the effectiveness of our method. Our simulations also demonstrated that 3D biometrics are better than 2D biometrics. Code is available at https://github.com/lhaof/LFC.

Radiomics Diagnosis of Microvascular Invasion in Hepatocellular Carcinoma Using 3D Ultrasound and Whole-Slide Image Fusion.

This study aims to develop a machine learning model that accurately diagnoses microvascular invasion (MVI) in hepatocellular carcinoma by using radiomic features from MVI-positive regions of interest (ROIs). Unlike previous studies, which do not account for the location and distribution of MVI, this research focuses on correlating preoperative imaging with postoperative pathological MVI. This study involves obtaining ex vivo 3D ultrasound images of 36 hepatic specimens from nine rabbits. These images are fused with whole-slide images to localize MVI regions precisely. The identified MVI regions are segmented into MVI-positive ROIs, with a 1:3 ratio of positive to negative ROIs. Radiomic features are extracted from each ROI, and 30 features highly associated with MVI are selected for model development. The performance of several machine learning models is evaluated using metrics such as sensitivity, specificity, accuracy, the area under the curve (AUC), and F1 score. The GBDT model achieves the best results, with an AUC of 0.91, an F1 score of 0.85, a sensitivity of 0.76, a specificity of 0.92, and an accuracy of 0.86. The high diagnostic accuracy of these models highlights the potential for future clinical application in the precise diagnosis of MVI using radiomic features from MVI-positive ROIs.

SwinDAF3D: Pyramid Swin Transformers with Deep Attentive Features for Automated Finger Joint Segmentation in 3D Ultrasound Images for Rheumatoid Arthritis Assessment

Rheumatoid arthritis (RA) is a chronic autoimmune disease that can cause severe joint damage and functional impairment. Ultrasound imaging has shown promise in providing real-time assessment of synovium inflammation associated with the early stages of RA. Accurate segmentation of the synovium region and quantification of inflammation-specific imaging biomarkers are crucial for assessing and grading RA. However, automatic segmentation of the synovium in 3D ultrasound is challenging due to ambiguous boundaries, variability in synovium shape, and inhomogeneous intensity distribution. In this work, we introduce a novel network architecture, Swin Transformers with Deep Attentive Features for 3D segmentation (SwinDAF3D), which integrates Swin Transformers into a Deep Attentive Features framework. The developed architecture leverages the hierarchical structure and shifted windows of Swin Transformers to capture rich, multi-scale and attentive contextual information, improving the modeling of long-range dependencies and spatial hierarchies in 3D ultrasound images. In a six-fold cross-validation study with 3D ultrasound images of RA patients’ finger joints (n = 72), our SwinDAF3D model achieved the highest performance with a Dice Score (DSC) of 0.838 ± 0.013, an Intersection over Union (IoU) of 0.719 ± 0.019, and Surface Dice Score (SDSC) of 0.852 ± 0.020, compared to 3D UNet (DSC: 0.742 ± 0.025; IoU: 0.589 ± 0.031; SDSC: 0.661 ± 0.029), DAF3D (DSC: 0.813 ± 0.017; IoU: 0.689 ± 0.022; SDSC: 0.817 ± 0.013), Swin UNETR (DSC: 0.808 ± 0.025; IoU: 0.678 ± 0.032; SDSC: 0.822 ± 0.039), UNETR++ (DSC: 0.810 ± 0.014; IoU: 0.684 ± 0.018; SDSC: 0.829 ± 0.027) and TransUNet (DSC: 0.818 ± 0.013; IoU: 0.692 ± 0.017; SDSC: 0.815 ± 0.016) models. This ablation study demonstrates the effectiveness of combining a Swin Transformers feature pyramid with a deep attention mechanism, improving the segmentation accuracy of the synovium in 3D ultrasound. This advancement shows great promise in enabling more efficient and standardized RA screening using ultrasound imaging.

Pharmacological inhibition of SIRT1 limits the growth of tumoral and metastatic granulosa cells by impacting mTOR, Myc and E2F pathways.

Clinical management of patients with ovarian granulosa cell tumor (GCT) remains poor. Sirtuin-1 (SIRT1), a deacetylase enzyme involved in the regulation of tumor growth and metastasis, may represent a therapeutic target due to the availability of selective pharmacological inhibitors with minimal toxicity. We assessed the possible overexpression of SIRT1 during tumorigenesis by Western blot and immunohistochemistry. We tested the effects of SIRT1 inhibition by EX-527 on growth, proliferation, death, migration, metabolism and gene expression by RNA sequencing in vitro on three GCT cell lines (AT29, KGN, COV434). Tumor growth in response to EX-527 treatment was examined in nude mice carrying subcutaneous GCT cell grafts using an electronic caliper and in GCT of AT83 mice by 3D ultrasound imaging system. SIRT1 abundance increased during tumorigenesis. In vitro treatment with EX-527 efficiently reduced cell growth, either by inducing apoptosis or by inhibiting proliferation. EX-527 induced alterations in mTOR, Myc and E2F-driven pathways, and in those controlling cell metabolism and oxidative stress. The administration of this treatment for 4 weeks efficiently reduced tumor progression in vivo. Inhibition of SIRT1 activity may have GCT-growth suppressive effects, providing a rationale for evaluating the therapeutic potential of drugs targeting SIRT1 in patients.

AI and early diagnostics: mapping fetal facial expressions through development, evolution, and 4D ultrasound.

The development of facial musculature and expressions in the human fetus represents a critical intersection of developmental biology, neurology, and evolutionary anthropology, offering insights into early neurological and social development. Fetal facial expressions, shaped by Cranial Nerve VII, reflect evolutionary adaptations for nonverbal communication and exhibit minimal asymmetry in universal expressions. Advancements in 4D ultrasound imaging and artificial intelligence (AI) have introduced innovative methods for analyzing these movements, revealing their potential as diagnostic tools for neurodevelopmental disorders like Bell's Palsy and Ramsay Hunt Syndrome before birth. These technologies promise early interventions that could significantly improve neonatal outcomes. By integrating imaging, AI, and longitudinal studies, researchers propose a multidisciplinary approach to establish diagnostic criteria for fetal facial movements. However, translating these advancements into clinical practice requires addressing ethical and practical challenges, refining imaging and AI methodologies, and fostering interdisciplinary collaboration. The review highlights the universality of fetal expressions while emphasizing the importance of distinguishing typical variability from pathological markers. In conclusion, these findings suggest transformative potential for maternal-fetal medicine, paving the way for proactive strategies to manage neurodevelopmental risks. Focused research is essential to fully harness these innovations and establish a new frontier in perinatal science.

Comparisons of Interapical Distance and Coronal Balance Measurements among Standing Positions in Participants with and without Adolescent Idiopathic Scoliosis Using 3D Ultrasound Imaging.

Cross-sectional study. The objective of this study was to quantify whether interapical distance and coronal balance measured on 3D ultrasound (3DUS) images differed among ten standing positions of participants with and without adolescent idiopathic scoliosis (AIS). Apical vertebral translation (AVT) is an indicator of clinical symptoms and treatment outcomes in AIS. Stereo-radiography simultaneously captures a frontal and lateral image, yet patients are required to elevate their arms during the examination and varied positioning may impact coronal plane measurements. Ultrasound can assess these measurements without exposing participants to radiation. Females with and without AIS were recruited from a scoliosis clinic, and email advertisements, respectively. Participants underwent 3DUS scans in ten positions: standing; arms anteriorly supported in 60° of shoulder flexion; fingers to clavicles, chin, zygomatic processes, and eyebrows; shoulders abducted at 90° with hands open and thumbs on shoulders; hands on anterior wall with and without blocks; and hands unsupported. AVT and coronal balance measurements were obtained using custom software. Positions and groups were compared using repeated measures ANOVAs with Sidak pairwise comparisons. Fifty-nine females had a mean age, height, and weight of 17.5±4.9 years, 162.8±5.8cm, and 56.2±10.6kg, respectively. Seventeen single and 14 double curve participants were included with mean maximum curve angles of 26.4°±3.6° and 25.2°±3.1°, respectively. Overall, there were no statistically significant differences in interapical distance or coronal balance among the ten positions, as well as in coronal balance between the groups (P>0.05). The interapical distance of participants without AIS significantly differed from those with single curves in eight positions and from those with double curves in all ten positions. The positions evaluated may be interchangeable for coronal plane lateral deviation measurements of females with AIS.

A new method for placental volume measurements using tracked 2D ultrasound and automatic image segmentation

Background Placental volume measurements can potentially identify high-risk pregnancies. We aimed to develop and validate a new method for placental volume measurements using tracked 2D ultrasound and automatic image segmentation. Methods We included 43 pregnancies at gestational week 27 and acquired placental images using a 2D ultrasound probe with position tracking, and trained a convolutional neural network (CNN) for automatic image segmentation. The automatically segmented 2D images were combined with tracking data to calculate placental volume. For 15 of the included pregnancies, placental volume was also estimated based on MRI examinations, 3D ultrasound and manually segmented 2D ultrasound images. The ultrasound methods were compared to MRI (gold standard). Results The CNN demonstrated good performance in automatic image segmentation (F1-score 0.84). The correlation with MRI-based placental volume was similar for tracked 2D ultrasound using automatically segmented images (absolute agreement intraclass correlation coefficient [ICC] 0.58, 95% CI 0.13–0.84) and manually segmented images (ICC 0.59, 95% CI 0.13–0.84). The 3D ultrasound method showed lower ICC (0.35, 95% CI −0.11 to 0.74) than the methods based on tracked 2D ultrasound. Conclusions Tracked 2D ultrasound with automatic image segmentation is a promising new method for placental volume measurements and has potential for further improvement.

Non-contact tracking of shoulder bones using ultrasound and stereophotogrammetry.

We explored the integration of 3D ultrasound (US) imaging with motion capture technology for non-invasively tracking bones of the shoulder district during normal activity. Our study aimed to demonstrate ex-vivo the proposed 3D US method's feasibility and accuracy of tracking shoulder bones in a controlled cadaveric shoulder positioned in various arm elevations (low, mid and high). We registered previously acquired full bone shapes to spatially small bony surface patches segmented from 3D US. The bone registration approach used was based on in silico analyses that investigated the effects of different - 1) registration algorithms (Iterative-Closest-Point-ICP, and Coherent Point Drift-CPD) and 2) initial estimate levels of the bone model pose relative to the targeted final bone pose-on the overall registration efficiency and accuracy in a controlled environment. CPD provided the highest accuracy in the simulation at the cost of 8x longer computation compared to ICP. The RMSE errors were <1mm for the humerus and scapula at all elevations. Ex-vivo, the CPD registration errors were (Humerus = 2mm and Scapula = 13.9mm) (Humerus = 7.2mm and Scapula = 16.8mm) and (Humerus = 14.28mm and Scapula = 27.5mm), for low, medium and high elevations respectively. In summary, we demonstrated the feasibility and accuracy of tracking shoulder bones with 3D US in a simulation and a cadaveric experiment. We discovered that CPD may be a more suitable registration method for the task than ICP. We also discussed that 3D US with motion capture technology is very promising for dynamic bone tracking, but the US technology may not be ready for the task yet.

Prior-Knowledge Embedded U-Net-Based Fully Automatic Vessel Wall Volume Measurement of the Carotid Artery in 3D Ultrasound Image.

The vessel-wall-volume (VWV) measured based on three-dimensional (3D) carotid artery (CA) ultrasound (US) images can help to assess carotid atherosclerosis and manage patients at risk for stroke. Manual involvement for measurement work is subjective and requires well-trained operators, and fully automatic measurement tools are not yet available. Thereby, we proposed a fully automatic VWV measurement framework (Auto-VWV) using a CA prior-knowledge embedded U-Net (CAP-UNet) to measure the VWV from 3D CA US images without manual intervention. The Auto-VWV framework is designed to improve the repeated VWV measuring consistency, which resulted in the first fully automatic framework for VWV measurement. CAP-UNet is developed to improve segmentation accuracy on the whole CA, which composed of a U-Net type backbone and three additional prior-knowledge learning modules. Specifically, a continuity learning module is used to learn the spatial continuity of the arteries in a sequence of image slices. A voxel evolution learning module was designed to learn the evolution of the artery in adjacent slices, and a topology learning module was used to learn the unique topology of the carotid artery. In two 3D CA US datasets, CAP-UNet architecture achieved state-of-the-art performance compared to eight competing models. Furthermore, CAP-UNet-based Auto-VWV achieved better accuracy and consistency than Auto-VWV based on competing models in the simulated repeated measurement. Finally, using 10 pairs of real repeatedly scanned samples, Auto-VWV achieved better VWV measurement reproducibility than intra- and inter-operator manual measurements. The code is available at https://github.com/Yue9603/Auto-VWV.

Deep Beamforming for Real-time 3D Passive Acoustic Mapping with Row-Column-Addressed Arrays.

Passive acoustic mapping (PAM) is a promising tool to monitor acoustic cavitation activities for focused ultrasound (FUS) therapies. While 2D matrix arrays allow three-dimensional (3D) PAM, the high channel count requirement and the complexity of the receiving electronics limit their practical value in real-time imaging applications. In this regard, row-column-addressed (RCA) arrays have shown great potentials in addressing the difficulties in real-time 3D ultrasound imaging. However, currently, there is no applicable method for 3D PAM with RCA arrays. In this work, we propose a deep beamformer for real-time 3D PAM with RCA arrays. The deep beamformer leverages a deep neural network to map radio-frequency microbubble (MB) cavitation signals acquired with the RCA array to 3D PAM images, achieving similar image quality to the reconstructions performed using the fully populated 2D matrix array with the angular spectrum method. In the simulation, the images reconstructed by the deep beamformer showed less than 13.2% and 1.8% differences in the energy spread volume and image signal-to-noise ratio, compared to those reconstructed using the matrix array. However, the image reconstruction time was reduced by 11 and 30 times on the CPU and GPU, respectively, achieving 42.4 volumes per second image reconstruction speed on a GPU for a volume sized 128×128×128. Experimental data further validated the capabilities of the deep beamformer to accurately localize MB cavitation activities in 3D space. These results clearly demonstrated the feasibility of real-time and 3D monitoring of MB cavitation activities with RCA arrays and neural network based beamformers.

Three-dimensional diffractive acoustic tomography

Acoustically probing biological tissues with light or sound, photoacoustic and ultrasound imaging can provide anatomical, functional, and/or molecular information at depths far beyond the optical diffusion limit. However, most photoacoustic and ultrasound imaging systems rely on linear-array transducers with elevational focusing and are limited to two-dimensional imaging with anisotropic resolutions. Here, we present three-dimensional diffractive acoustic tomography (3D-DAT), which uses an off-the-shelf linear-array transducer with single-slit acoustic diffraction. Without jeopardizing its accessibility by general users, 3D-DAT has achieved simultaneous 3D photoacoustic and ultrasound imaging with optimal imaging performance in deep tissues, providing near-isotropic resolutions, high imaging speed, and a large field-of-view, as well as enhanced quantitative accuracy and detection sensitivity. Moreover, powered by the fast focal line volumetric reconstruction, 3D-DAT has achieved 50-fold faster reconstruction times than traditional photoacoustic imaging reconstruction. Using 3D-DAT on small animal models, we mapped the distribution of the biliverdin-binding serpin complex in glassfrogs, tracked gold nanoparticle accumulation in a mouse tumor model, imaged genetically-encoded photoswitchable tumors, and investigated polyfluoroalkyl substances exposure on developing embryos. With its enhanced imaging performance and high accessibility, 3D-DAT may find broad applications in fundamental life sciences and biomedical research.

A Force-Sensor-Less Approach for Rapid Young's Modulus Identification of Heterogeneous Soft Tissue.

Due to individual differences, accurate identification of tissue elastic parameters is essential for biomechanical modeling in surgical guidance for hepatic venous injections. This paper aims to acquire the absolute Young's modulus of heterogeneous soft tissues during endoscopic surgery with 2D ultrasound images. First, we introduced a force-sensor-less approach that utilizes a pre-calibrated soft patch with a known Young's modulus and its ultrasound images to calculate the external forces exerted by the probe on the tissue. Second, we introduced a Kriging-based inverse algorithm to identify the relative Young's modulus (RYM) between the inclusion and the background tissue. The RYM was estimated based on 2D plane strain approximation and mapped to the RYM of 3D soft tissue through a trained Kriging model. Finally, we developed a direct method to identify the background Young's modulus (BYM) based on calculated external forces and RYM. The simulation results demonstrate the high efficiency and robustness of the Kriging-based inverse algorithm in identifying RYM. Physical experiments on the three phantoms show that the errors of the identified BYM and RYM are all below 15%. The proposed methodology for Young's modulus identification is feasible and achieves satisfactory accuracy and computational efficiency in both simulations and physical experiments.

Three-dimensional complementary breast ultrasound (3D CBUS): Improving 3D spatial resolution uniformity with orthogonal images.

With increasing evidence supporting three-dimensional (3D) automated breast (AB) ultrasound (US) for supplemental screening of breast cancer in increased-risk populations, including those with dense breasts and in limited-resource settings, there is an interest in developing more robust, cost-effective, and high-resolution 3DUS imaging techniques. Compared with specialized ABUS systems, our previously developed point-of-care 3D ABUS system addresses these needs and is compatible with any conventional US transducer, which offers a cost-effective solution and improved availability in clinical practice. While conventional US transducers have high in-plane resolution (axial and lateral), their out-of-plane resolution is constrained by the poor intrinsic elevational US resolution. Consequently, any oblique view plane in an acquired 3DUS image will contain high in-plane and poor out-of-plane resolution components, diminishing spatial resolution uniformity and overall diagnostic utility. To develop and validate a novel 3D complementary breast ultrasound (CBUS) approach to improve 3DUS spatial resolution uniformity using a conventional US transducer by acquiring and generating orthogonal 3DUS images. We previously developed a cost-effective, portable, dedicated 3D ABUS system consisting of a wearable base, a compression assembly, and a mechanically driven scanner for automated 3DUS image acquisition, compatible with any commercial linear US transducer. For this system, we have proposed 3D CBUS approach which involves acquiring and registering orthogonal 3DUS images ( and ) with an aim of overcoming the poor resolution uniformity in the scanning direction in 3D US images. The voxel intensity values in the 3D CBUS image are computed with a spherical-weighted algorithm from the original orthogonal 3DUS images. Experimental validation was performed with 2DUS frame densities of 2, 4, 6 frames mm-1 using an agar-based phantom with a speed of sound of 1540 ms-1 and an embedded nylon bead. Lateral and axial full-width at half-maximum (FWHMLAT and FWHMAX) values were calculated from cross-sections taken at polar view planes ranging from 0° to 90° for 3DUS and 3D CBUS images of a bead phantom in focal zone and far field regions. Kendall's Tau-b correlation coefficients were calculated between FWHM measurements and cross-section angle for all frame density settings at a significance level of . Volumetric 3D segmentations were performed for 3DUS and 3D CBUS images of an inclusion phantom to confirm volumetric reconstruction accuracy. For statistical analysis, a repeated measures ANOVA with the Greenhouse-Geisser correction was performed at a significance level of . Experimental validation of the orthogonal 3DUS images show complementary trends of increasing and decreasing FWHMLAT from in-plane to out-of-plane (0° and 90° and vice versa) views. This is exemplified with the scan taken at 4 frames mm-1 in the focal zone, where FWHMLAT ranges from 3.51 to 1.10 mm for and 1.02-3.02 mm for , spanning 0°-90°, respectively. When combined in the 3D CBUS image, the FWHMLAT exhibits greater uniformity across view angles by mitigating poor out-of-plane resolution using its complementary in-plane component, with corresponding FWHMLAT values of 1.27 and 1.46 mm. While visual enhancements were seen in the 3D CBUS image, no statistically significant differences were found in volumetric measurements of the spherical inclusions in the 3DUS and 3D CBUS images. The out-of-plane resolution in the orthogonal 3DUS images is improved upon their combination into a single 3D CBUS image. These results demonstrate that the proposed 3D CBUS generation approach can improve 3D spatial resolution uniformity, while employing a commercial US transducer. The proposed 3D CBUS method shows potential utility for improving image resolution uniformity in 3D ABUS images, with the goal of improving point-of-care breast cancer supplemental screening and diagnostic applications, particularly in women with dense breasts and limited resource settings.

Automatic Measurement of Frontomaxillary Facial Angle in Fetal Ultrasound Images Using Deep Learning.

Accurate measurement of frontomaxillary facial (FMF) angles in prenatal ultrasound (US) scans plays a pivotal role in the screening of trisomy 21. Nevertheless, this intricate procedure heavily relies on the proficiency of the ultrasonographer and tends to be a time-intensive task. Furthermore, FMF angles are subjective when measured manually. To address this challenge, we propose a deep learning-based assisted examination framework for automatically measuring FMF angles on 2D ultrasound images. Firstly, we trained a deep learning network using 1549 fetal ultrasound images to achieve automatic and accurate segmentation of critical areas. Subsequently, a key point detection network was employed to predict the coordinates of the requisite points for calculating FMF angles. Finally, FMF angles were obtained through computational means. We employed Pearson correlation coefficients and Bland-Altman plots to assess the correlation and consistency between the model's predictions and manual measurements. Notably, our method exhibited a mean absolute error of 2.354°, outperforming the typical standards of the junior expert. This indicates the high degree of accuracy and reliability achieved by our approach.

How Reproducible Are the Ultrasound Features of Adenomyosis Defined by the Revised MUSA Consensus?

Background/Objectives: The aim of this study is to assess the inter- and intra-observer reproducibility of the identification of direct and indirect ultrasonographic features of adenomyosis as defined by the revised Morphological Uterus Sonographic Assessment (MUSA) consensus (2022). Methods: A cohort of 74 women, aged 18 to 45, were recruited from the recurrent miscarriage and general gynaecology clinic at a university-based fertility centre. All the participants underwent 2D and 3D transvaginal Ultrasound scan (TVS) examination in the late follicular and early luteal phase. Conventional grey scale and power Doppler image volumes were acquired and stored. Subsequently, the stored 3D ultrasound images were independently re-evaluated offline by the two observers for the direct and indirect features of adenomyosis as outlined by the revised MUSA group. The intra- and the inter-observer reproducibility was estimated using Cohen's Kappa coefficient. Results: The intra- and interobserver reproducibility (K -0.27, 95% CI 0.06-0.48 and K 0.13, 95% CI -0.10-0.37, respectively) for at least one direct feature of adenomyosis was only modest. Amongst the individual direct features, the interobserver variability of identifying myometrial cysts was fair (K 0.21, 95% CI -0.00-0.42), whereas the intra-observer variability was moderate (K 0.44, 95% CI 0.26-0.63). While hyperechogenic islands identification achieved a fair level of intra- (K 0.31, 95% CI 0.09-0.53) and interobserver (K 0.24, 95% CI 0.01-0.47) agreement, the reproducibility of reporting sub-endometrial lines/buds was fair for the intra-observer (K 0.22, 95% CI -0.02 0.47) and poor for the interobserver (K 0.00, 95% CI -0.20-0.19). The interobserver agreement for indirect features varied from poor to moderate, while the intra-observer agreement ranged between poor to good. Conclusions: The reporting of adenomyosis using direct features suggested by the revised MUSA group consensus showed only modest interobserver and intra-observer agreement. The definitions of ultrasound features for adenomyosis need further refining to enhance the reliability of diagnosis criteria of adenomyosis.

The effect of arm positions used during radiography on spinal alignment parameters assessed by 3D ultrasound imaging in adolescents with and volunteers without idiopathic scoliosis

PurposeClinicians monitor scoliosis progression using multiple radiographs during growth. During imaging, arms must be elevated to visualize vertebrae, possibly affecting sagittal alignment. This study aimed to determine the arm position that best represents habitual standing (and possibly allowing hand-based skeletal maturity assessment) to obtain frontal and lateral stereo-radiographs as measured using frontal, sagittal, and transverse angles.MethodsFemales with and without, and males with Adolescent Idiopathic Scoliosis (AIS) were recruited consecutively. Patients were scanned using 3D Ultrasound imaging (3DUS), in 10 arm positions; habitual standing, arms supported anteriorly at 60° flexion, fingers to clavicle, chin, zygomatic, and eyebrows, arms abducted 90°, hands on wall, on blocks, and unsupported. Axial vertebral rotation (AVR) differences, frontal, and sagittal curve angles were measured. Repeated measures ANOVAs with Sidak post-hoc tests compared positions.ResultsNinety females with and without AIS with mean age, and height of 17 ± 4 years, and 162 ± 6 cm, and ten males with AIS of 16 ± 3 years, and 174 ± 11 cm, respectively, were included. Female AIS single-curve showed larger curves in standing in all positions excluding hands on blocks (p > 0.05). Sagittal parameters showed decreases in kyphosis in arms abducted 90° and increases in lordosis in fingers to cheeks/eyebrows (p > 0.05). AVR twist was not significantly affected by position. Male AIS showed comparable results to females, but no significant differences were detected.ConclusionNo position represented habitual standing for all groups. When arms are raised, decreases in curve angle were shown in single-curve patients, kyphosis decreased, and lordosis increased in all groups. Most accurate positioning for all parameters was in fingers to clavicle/chin position.

Multimodal Study of Murine Cardiovascular Remodeling: Four-Dimensional Ultrasound and Mass Spectrometry Imaging.

Cardiovascular disease (CVD) is the leading cause of death in the United States. Damage inthe cardiovascular system can be due to environmental exposure, trauma, drug toxicity, or numerous other factors.As a result, cardiac tissue and vasculature undergo structural changes and display diminished function. The damage and the resulting remodeling can be detected and quantified with ultrasound (US) imaging at the organ level and mass spectrometry imaging (MSI) at the molecular level. This manuscript describes an innovative methodology for studying murine cardiac pathophysiology, coupling in vivo four-dimensional (4D) ultrasound imaging and analysis with ex vivo matrix-assisted laser desorption/ionization (MADLI) MSI of the heart. 4D ultrasound can provide dynamic volumetric measurements, including radial displacement, surface area strain, and longitudinal strain throughout an entire cardiac cycle. In the vasculature, MSI and ultrasound are used to assess vessel wall compositions, hemodynamics, and vessel wall dynamics. The methodology can be tailored to study a myriad of CV diseases by adjusting functional metrics of interest and/or varying MALDI MSI protocol to target specific molecules. MALDI MSI can be used to study lipids, small metabolites, peptides, and glycans. This protocol outlines the use of MALDI MSI for untargeted lipidomic analysis and the use of ultrasound imaging for cardiovascular hemodynamics and biomechanics.