Ultrasound Video Research Articles

Sensorless volumetric reconstruction of fetal brain freehand ultrasound scans with deep implicit representation

Three-dimensional (3D) ultrasound imaging has contributed to our understanding of fetal developmental processes by providing rich contextual information of the inherently 3D anatomies. However, its use is limited in clinical settings, due to the high purchasing costs and limited diagnostic practicality. Freehand 2D ultrasound imaging, in contrast, is routinely used in standard obstetric exams, but inherently lacks a 3D representation of the anatomies, which limits its potential for more advanced assessment. Such full representations are challenging to recover even with external tracking devices due to internal fetal movement which is independent from the operator-led trajectory of the probe. Capitalizing on the flexibility offered by freehand 2D ultrasound acquisition, we propose ImplicitVol to reconstruct 3D volumes from non-sensor-tracked 2D ultrasound sweeps. Conventionally, reconstructions are performed on a discrete voxel grid. We, however, employ a deep neural network to represent, for the first time, the reconstructed volume as an implicit function. Specifically, ImplicitVol takes a set of 2D images as input, predicts their locations in 3D space, jointly refines the inferred locations, and learns a full volumetric reconstruction. When testing natively-acquired and volume-sampled 2D ultrasound video sequences collected from different manufacturers, the 3D volumes reconstructed by ImplicitVol show significantly better visual and semantic quality than the existing interpolation-based reconstruction approaches. The inherent continuity of implicit representation also enables ImplicitVol to reconstruct the volume to arbitrarily high resolutions. As formulated, ImplicitVol has the potential to integrate seamlessly into the clinical workflow, while providing richer information for diagnosis and evaluation of the developing brain.

Establishment and preliminary application of personalized three-dimensional reconstruction of thyroid gland with automatic detection of thyroid nodules based on ultrasound videos.

A well display of the spatial location of thyroid nodules in the thyroid is important for surgical path planning and surgeon-patient communication. The aim of this study was to establish a three-dimensional (3D) reconstruction method of the thyroid gland, thyroid nodule, and carotid artery with automatic detection based on two-dimensional (2D) ultrasound videos, and to evaluate its clinical value. Ultrasound videos, including the thyroid gland with nodule, isthmus of thyroid gland, and ipsilateral carotid artery, were recorded. BC-UNet, MTN-Net, and RDPA-U-Net network models were innovatively employed for segmentation of the thyroid glands, the thyroid nodules, and the carotid artery respectively. Marching Cubes algorithm was used for reconstruction, while Laplacian smoothing algorithm was employed to smooth the 3D model surface. Using this model, 20 patients and 15 surgeons completed surveys on the effectiveness of this model for the pre-surgery demonstration of nodule location as well as surgeon-patient communication. The thyroid gland with nodule, isthmus of gland, and carotid artery were reconstructed and displayed. With the 3D model, the understanding of the spatial location of thyroid nodules improved in all three surgeon groups, eliminating the influence of professional levels. In the patient survey, the patients' understanding of the thyroid nodule location and procedure for surgery were significantly improved. In addition, with the 3D model, the time for doctors to explain to patients was significantly reduced (16.75 vs. 8.85min, p=0.001). To our knowledge, this is the first report of constructing a 3D thyroid model using a deep learning technique for personalized thyroid segmentation based on 2D ultrasound videos. The preliminary clinical application showed that it was conducive to the comprehension of the location of thyroid nodules for surgeons and patients, with significant improvement on the efficiency of surgeon-patient communication.

Long-short diffeomorphism memory network for weakly-supervised ultrasound landmark tracking

Ultrasound is a promising medical imaging modality benefiting from low-cost and real-time acquisition. Accurate tracking of an anatomical landmark has been of high interest for various clinical workflows such as minimally invasive surgery and ultrasound-guided radiation therapy. However, tracking an anatomical landmark accurately in ultrasound video is very challenging, due to landmark deformation, visual ambiguity and partial observation. In this paper, we propose a long-short diffeomorphism memory network (LSDM), which is a multi-task framework with an auxiliary learnable deformation prior to supporting accurate landmark tracking. Specifically, we design a novel diffeomorphic representation, which contains both long and short temporal information stored in separate memory banks for delineating motion margins and reducing cumulative errors. We further propose an expectation maximization memory alignment (EMMA) algorithm to iteratively optimize both the long and short deformation memory, updating the memory queue for mitigating local anatomical ambiguity. The proposed multi-task system can be trained in a weakly-supervised manner, which only requires few landmark annotations for tracking and zero annotation for deformation learning. We conduct extensive experiments on both public and private ultrasound landmark tracking datasets. Experimental results show that LSDM can achieve better or competitive landmark tracking performance with a strong generalization capability across different scanner types and different ultrasound modalities, compared with other state-of-the-art methods.

Noninvasive Assessment of Intracranial Pressure: Deformability Index as an Adjunct to Optic Nerve Sheath Diameter to Increase Diagnostic Ability

BackgroundToday, invasive intracranial pressure (ICP) measurement remains the standard, but its invasiveness limits availability. Here, we evaluate a novel ultrasound-based optic nerve sheath parameter called the deformability index (DI) and its ability to assess ICP noninvasively. Furthermore, we ask whether combining DI with optic nerve sheath diameter (ONSD), a more established parameter, results in increased diagnostic ability, as compared to using ONSD alone.MethodsWe prospectively included adult patients with traumatic brain injury with invasive ICP monitoring, which served as the reference measurement. Ultrasound images and videos of the optic nerve sheath were acquired. ONSD was measured at the bedside, whereas DI was calculated by semiautomated postprocessing of ultrasound videos. Correlations of ONSD and DI to ICP were explored, and a linear regression model combining ONSD and DI was compared to a linear regression model using ONSD alone. Ability of the noninvasive parameters to distinguish dichotomized ICP was evaluated using receiver operating characteristic curves, and a logistic regression model combining ONSD and DI was compared to a logistic regression model using ONSD alone.ResultsForty-four ultrasound examinations were performed in 26 patients. Both DI (R = − 0.28; 95% confidence interval [CI] R < − 0.03; p = 0.03) and ONSD (R = 0.45; 95% CI R > 0.23; p < 0.01) correlated with ICP. When including both parameters in a combined model, the estimated correlation coefficient increased (R = 0.51; 95% CI R > 0.30; p < 0.01), compared to using ONSD alone, but the model improvement did not reach statistical significance (p = 0.09). Both DI (area under the curve [AUC] 0.69, 95% CI 0.53–0.83) and ONSD (AUC 0.72, 95% CI 0.56–0.86) displayed ability to distinguish ICP dichotomized at ICP ≥ 15 mm Hg. When using both parameters in a combined model, AUC increased (0.80, 95% CI 0.63–0.90), and the model improvement was statistically significant (p = 0.02).ConclusionsCombining ONSD with DI holds the potential of increasing the ability of optic nerve sheath parameters in the noninvasive assessment of ICP, compared to using ONSD alone, and further study of DI is warranted.

Hyoid Bone Movement During Swallowing in Female Thyroidectomy Patients: A Kinematic Ultrasound Study.

The aim of this study was to investigate the measures of displacement, time and velocity of hyoid bone movement in female thyroidectomy patients. Fifty-eight ultrasound videos of 29 women during swallowing were analyzed. The sample was divided into experimental group (EG), composed of 12 women following total or partial thyroidectomy; and control group (CG) of 17 healthy women. The kinematic measures of displacement, time and velocity of hyoid bone displacement were tracked during swallowing of 10ml of liquid (water) and 10ml of thickened liquid (honey) in both groups for comparisons. Additional analysis included bolus consistency effect and relationship with clinical characteristics. Ultrasound videos were analyzed according to a standardized protocol using ImageJ software. Displacement, time and velocity of hyoid movement during swallowing of 10ml of liquid or thickened liquid were not statistically different between female thyroidectomy patients and healthy women. There is no bolus consistency effect on kinematic measures in both groups, but among thyroidectomy patients, velocity of hyoid bone movement is significantly faster in those with swallowing complaints.

Automating the Human Action of First-Trimester Biometry Measurement from Real-World Freehand Ultrasound

Objective: Automated medical image analysis solutions should closely mimic complete human actions to be useful in clinical practice. However, more often an automated image analysis solution only represents part of a human task which restricts its practical utility. In the case of ultrasound-based fetal biometry, an automated solution should ideally recognize key fetal structures in free-hand video guidance, select a standard plane from a video stream and perform biometry. A complete automated solution should automate all three sub-actions.Methods: This paper considers how to automate the complete human action of first-trimester biometry measurement from real-world freehand ultrasound. In the proposed hybrid Convolutional Neural Network (CNN) architecture design, a classification-regression-based guidance model detects and tracks fetal anatomical structures (using visual cues) in the ultrasound video. Several high-quality standard planes which contain the mid-sagittal view of the fetus are sampled at multiple timestamps (using a custom-designed confident-frame detector) based on the estimated probability values associated with predicted anatomical structures that define the biometry plane. Automated semantic segmentation is performed on the selected frames to extract fetal anatomical landmarks. A crown-rump length (CRL) estimate is calculated as the mean CRL from these multiple frames.Results: Our fully automated method shows a high correlation with clinical expert CRL measurement (Pearson ρ=0.92, R-Squared (R2)=0.84) and a low mean absolute error of 0.834 (weeks) for fetal age estimation on a test data set of 42 videos.Conclusion: A novel algorithm for standard plane detection employs a quality detection mechanism defined by clinical standards and ensuring precise biometric measurements.

Quantitative evaluation of adhesion severity around subscapularis and its relationship with shoulder range of motion in frozen shoulder and rotator cuff disorder: an observational study using dynamic ultrasonography

BackgroundThis study aimed to evaluate the severity of adhesion between muscles in the shoulder joint using dynamic ultrasonography and to confirm whether adhesions cause range of motion (ROM) restrictions. MethodsTwenty-four shoulders from fifteen frozen shoulder (FS) patients and Twenty-four shoulders from eighteen rotator cuff disorder (RCD) patients were enrolled. We obtained ultrasound video sequences of the subscapularis and deltoid muscles during shoulder external rotation. The mean stretching velocities of the deltoid (VDelt) and subscapularis (VSSC) were subsequently analyzed using a PC. If adhesions occurred between both muscles, the deltoid was stretched more vigorously, and we calculated VDelt/VSSC as adhesion severity. The coracohumeral ligament (CHL) thickness was measured using the same images. Shoulder ROM was measured by using a universal goniometer. ResultsThe intraclass correlation coefficients (1.1) and (2.1) of the adhesion severity measurements were 0.85 and 0.91, respectively. Multiple linear regression analysis revealed that the adhesion severity is a significant predictor for external rotation ROM in the RCD group (R2 = 0.44, F = 10.1, p < 0.01, t = -2.9), while CHL thickness predicts ROM in the FS group (R2 = 0.28, F = 5.5, p = 0.01, t = -3.0). ConclusionThe proposed method is reliable. Muscle adhesion causes ROM restriction of the shoulder joint. The primary cause of shoulder ROM restriction differed between the diagnostic groups.

Atrial Septal Defect Detection in Children Based on Ultrasound Video Using Multiple Instances Learning.

Thoracic echocardiography (TTE) can provide sufficient cardiac structure information, evaluate hemodynamics and cardiac function, and is an effective method for atrial septal defect (ASD) examination. This paper aims to study a deep learning method based on cardiac ultrasound video to assist in ASD diagnosis. We chose four standard views in pediatric cardiac ultrasound to identify atrial septal defects; the four standard views were as follows: subcostal sagittal view of the atrium septum (subSAS), apical four-chamber view (A4C), the low parasternal four-chamber view (LPS4C), and parasternal short-axis view of large artery (PSAX). We enlist data from 300 children patients as part of a double-blind experiment for five-fold cross-validation to verify the performance of our model. In addition, data from 30 children patients (15 positives and 15 negatives) are collected for clinician testing and compared to our model test results (these 30 samples do not participate in model training). In our model, we present a block random selection, maximal agreement decision, and frame sampling strategy for training and testing respectively, resNet18 and r3D networks are used to extract the frame features and aggregate them to build a rich video-level representation. We validate our model using our private dataset by five cross-validation. For ASD detection, we achieve AUC, accuracy, sensitivity, specificity, and F1 score. The proposed model is a multiple instances learning-based deep learning model for video atrial septal defect detection which effectively improves ASD detection accuracy when compared to the performances of previous networks and clinical doctors.

Utility of tele-guidance for point-of-care ultrasound: a single center prospective diagnostic study.

Point-of-care-ultrasound (POCUS) improves diagnostic accuracy and expedites lifesaving procedures. Remote areas disproportionately underuse ultrasound (US) due to a dearth of US trained professionals, imaging tools, and lack of quality assurance. Handheld US probes have been approved for diagnostic imaging but there have been limited studies examining their tele-guidance features. This study investigates whether physician tele-guidance improved ultrasound image acquisition by untrained scanners. To determine the effectiveness of tele-guidance for ultrasound image acquisition, 30 participants (15 in each study arm: experimental and control) with no ultrasound or medical experience gathered images of the heart, right kidney, and gallbladder of a standardized patient using a handheld ultrasound probe (Butterfly iQ +). All participants watched a standardized ultrasound tutorial video and were randomized into the control or experimental group. A physician assisted the experimental group using ultrasound probe's tele-guidance feature while the control group received no assistance. Time to image acquisition was recorded for both groups, and the images were graded by 3 blinded radiologists using the RACE tool to determine image and diagnostic quality. There was evidence that mean imaging time was greater in the control group for the heart, right kidney, and gallbladder (p < 0.0001, all; Cohen's DL: 2.0, 1.7, 3.0, respectively). Similarly, there was evidence that the predicted mean image quality for the heart, right kidney, and gallbladder was higher for the experimental group compared to the control group (3.46 versus 1.86, 4.49 versus 1.58, and 3.93 versus 1.5, respectively; p < 0.0001, all). There was also evidence that the diagnostic quality of images had a statistically higher predicted probability of meaningful interpretation for the experimental group for pericardial fluid, intraperitoneal fluid, and acute cholecystitis (p = 0.003, p < 0.0001, p < 0.0001, respectively). Tele-guidance improves time to image acquisition and clinical applicability of ultrasound images obtained by untrained scanners.

Ultrasound image quality of the carpal tunnel during a pinch grip: effects of coupling media and wrist position

ABSTRACT High quality ultrasound is required for evaluating carpal tunnel dynamics. The purpose of this study was to subjectively assess ultrasound video quality of the carpal tunnel with two commonly used coupling media (coupling gel, 7 mm standoff pad) in two wrist positions (0° neutral, 30° flexion) over the entire time course of a pinch grip. Seventeen participants completed a pinch grip while their carpal tunnel was scanned with ultrasound. The ultrasound videos were graded by two independent reviewers using 10-point scales encompassing median nerve detail, resolution, and total video quality. Intra- and inter-rater reliability were highest for median nerve detail (ICCs>.85). Ultrasound videos in the 0° neutral wrist condition resulted in significantly greater ratings of nerve detail compared to those in 30° wrist flexion. Future studies may benefit from a more customised approach to standoff pad thickness during dedicated analysis of the median nerve in a neutral wrist position.

(156) Morphometrix Tissue Segmentation Technology as a Diagnostic Aid in the Evaluation of Vasculogenic Causes of Erectile Dysfunction

Abstract Introduction Morphometrix Imaging Technologies Inc. (MPX) Tissue Segmentation Technology (TST) isolates and processes the component of ultrasound reflective signals associated with coherent specular reflection at the liquid spaces in living tissue. The software identifies and quantifies the hypoechoic structures in B-Mode Ultrasound Scans produced by these reflections and thereby profiles liquid spaces in the tissue created by blood and lymphatic vessels as a proxy for vascular intensity. It then derives tissue segment counts (TSC) as a final quantification step. Unlike SHIM Scores, TST provides a quantitative assessment of the vascular health of the erectile tissue. Objective The goal of the study was to determine whether TSC can predict a response to regenerative treatments for ED. We retrospectively studied the efficacy of MPX TST in assessing the vascularity of the corpus cavernosa (CC) and compared the findings to SHIM scores before and after Shockwave therapy (SWT). MPX can assess penile vascularity in the flaccid state using TSC and can calculate cross-sectional area and volume. Methods Retrospective study of 57 men, average age 54.9 with ED. B-mode Ultrasound video scans were performed on the penis in the flaccid state before and after treatment completion and processed through MPX to obtain the TSC which were compared to IIEF-5 SHIM scores. A typical treatment involved multiple sessions of SWT and PRP penile injection. Subjects were divided into three groups based on response to treatment and correlation with SHIM Scores: Group 1 – Good Responders &amp; Good Correlation with SHIMs, Group 2 – Poor &amp; Non-Responders &amp; Good Correlation with SHIMs, Group 3 – Good &amp; Poor Responders &amp; Non-Correlation with SHIMs. Results Overall we had an 80.7% correlation between TSC and the SHIM Score results. The average SHIM score increased to 15.9 from 10.6, (p&amp;lt;0.006). The average TSC in the left CC increased to 25 from 17.3, (p&amp;lt;0.002). The average TSC in the right CC increased to 24.7 from 17.9, (p&amp;lt;0.002). 54.9% (34 patients) were good responders and the SHIM scores correlated well with TSC. The average SHIM score increased to 17.5 from 10.8, (p&amp;lt;0.001). The average TSC in the left CC increased to 28.1 from 15.3, (p&amp;lt;0.001). The average TSC in the right CC increased to 31.2 from 17.1, (p&amp;lt;0.003). 21.1% (12 patients) were poor responders and the SHIM scores correlated well with TSC. The average SHIM score increased 10.3 from 9.4, (p&amp;lt;0.554). The average TSC in the left CC declined to 13.5 from 15.5, (p&amp;lt;0.559). The average TSC in the right CC declined slightly to 14.9 from 12.5, (p&amp;lt;0.268). 19.3% (11 patients) were a mixture of good and poor responders and the SHIM scores did not correlate well with TSC. The average SHIM score increased to 15.1 from 11.3, (p&amp;lt;0.103). The average TSC in the left CC declined to 25.5 from 26.8, (p&amp;lt;0.919). The average TSC in the right CC declined to 25.5 from 25.7, (p&amp;lt;0.975). Conclusions MPX tissue segmentation software can provide quantitative information about the vascularity of the corpus cavernosa and can differentiate responders to treatment from non-responders. Disclosure Any of the authors act as a consultant, employee or shareholder of an industry for: Morphometrix.

(126) Application of HIFEM Therapy Improves Orgasmic, Urinary and Erectile Function by Stimulation of Pelvic Floor Muscles

Abstract Introduction The weak pelvic floor in men contributes to various urological disorders. Treatment can be accomplished by strengthening the pelvic floor muscles (PFM) through pelvic floor muscle training. Nevertheless, such development of PFM is time-consuming and heavily dependent on the subject's ability to perform selective muscle contractions. An advantageous approach is the targeted stimulation of PFM which can be accomplished via electromagnetic stimulation. Particularly, the pelvic floor muscles can be non-invasively and selectively stimulated via the HIFEM field. This results in the natural development and rehabilitation of PFM, which, among other functions, plays a vital role in facilitating orgasmic, urinary and erectile function. Objective The aim of this study is to assess the safety and efficacy of the HIFEM pelvic procedure for the treatment of erectile dysfunction. Methods In this study, 28 males with a median age of 56 (30-72 years) were enrolled. The treatment was composed of ten weekly 28-minute sessions. Outcome measurements were collected via questionnaires and ultrasound imaging prior to and after the last treatment. To assess changes in erectile and sexual function, and prostate-related symptoms, the International Index of Erectile Function (IIEF) and International Prostate Symptom Score (IPSS) were used. Transverse B-Mode ultrasound video scans were recorded along the long axis of the penile shaft from the tip to the base with the penis in a flaccid state. A linear transducer was used, the depth was set at 2 cm, and the frame rate was approximately 30 frames per second. Separate video recordings of the left and right cavernosum were taken. The ultrasound scans were processed by the Morphometrix cloud-based algorithm. Adverse events were monitored during the study. Results All evaluation methods proved the efficiency of treatment on a statistically significant level. A significant improvement was detected in the majority (4 of 5) of the domains of IIEF, with the overall score improvement from a baseline value of 45.14±2.42 to 48.75±2.86 (+3.61 points). The most exceptional improvement was found in Sexual desire (+18.78%; p=1.06x10-3), Erectile dysfunction (+13.03%; p=5.7x10-4), and Orgasmic function (+10.69%; p=7.04x10-5) domains. The mean difference in the total score of the IPSS questionnaire reached -3.04 points (p=5.1x10-6), exceeding the corresponding minimal clinical difference (MIC=3 points). Overall, subjects, on average, shifted from moderately symptomatic (9.57±1.37 points) to mildly symptomatic (6.54±0.91 points). An increase in vascularity of both cavernosal was measured within 57% of patients who underwent ultrasound examination; specifically, a significant improvement was recorded in tissue segment count (TSC; +104%; p=1.17x10-4) and density (TSD; +81% p=2.20x10-3), pointing to better blood flow and delivery in examined tissues. No adverse event was observed. Conclusions The therapy administering HIFEM technology of the pelvic floor for selective PFM stimulation demonstrated significant improvement of erectile and orgasmic function and urinary symptoms. The questionnaire results corresponded with ultrasound measurements using a novel algorithm, identifying positive changes in the penile vascular system. Overall, the treatment may be considered a safe and a possible first-line solution for male patients suffering from dysfunctions linked to weakened pelvic floor muscles. Further investigation is needed. Disclosure Yes, this is sponsored by industry/sponsor: BTL. Clarification: Industry funding only - investigator initiated and executed study. Any of the authors act as a consultant, employee or shareholder of an industry for: AFFIRM Science.

Transformer-Based Automated Segmentation of the Median Nerve in Ultrasound Videos of Wrist-to-Elbow Region.

Segmenting the median nerve is essential for identifying nerve entrapment syndromes, guiding surgical planning and interventions, and furthering understanding of nerve anatomy. This study aims to develop an automated tool that can assist clinicians in localizing and segmenting the median nerve from the wrist, mid-forearm, and elbow in ultrasound videos. This is the first fully automated single deep learning model for accurate segmentation of the median nerve from the wrist to the elbow in ultrasound videos, along with the computation of the cross-sectional area (CSA) of the nerve. The visual transformer architecture, which was originally proposed to detect and classify 41 classes in YouTube videos, was modified to predict the median nerve in every frame of ultrasound videos. This is achieved by modifying the bounding box sequence matching block of the visual transformer. The median nerve segmentation is a binary class prediction, and the entire bipartite matching sequence is eliminated, enabling a direct comparison of the prediction with expert annotation in a frame-by-frame fashion. Model training, validation, and testing were performed on a dataset comprising ultrasound videos collected from 100 subjects, which were partitioned into 80, ten, and ten subjects, respectively. The proposed model was compared with U-Net, U-Net++, Siam U-Net, Attention U-Net, LSTM U-Net, and Trans U-Net. The proposed transformer-based model effectively leveraged the temporal and spatial information present in ultrasound video frames and efficiently segmented the median nerve with an average dice similarity coefficient (DSC) of approximately 94% at the wrist and 84% in the entire forearm region.

Cascaded Inner-Outer Clip Retformer for Ultrasound Video Object Segmentation.

Computer-aided ultrasound (US) imaging is an important prerequisite for early clinical diagnosis and treatment. Due to the harsh ultrasound (US) image quality and the blurry tumor area, recent memory-based video object segmentation models (VOS) achieve frame-level segmentation by performing intensive similarity matching among the past frames which could inevitably result in computational redundancy. Furthermore, the current attention mechanism utilized in recent models only allocates the same attention level among whole spatial-temporal memory features without making distinctions, which may result in accuracy degradation. In this paper, we first build a larger annotated benchmark dataset for breast lesion segmentation in ultrasound videos, then we propose a lightweight clip-level VOS framework for achieving higher segmentation accuracy while maintaining the speed. The Inner-Outer Clip Retformer is proposed to extract spatialtemporal tumor features in parallel. Specifically, the proposed Outer Clip Retformer extracts the tumor movement feature from past video clips to locate the current clip tumor position, while the Inner Clip Retformer detailedly extracts current tumor features that can produce more accurate segmentation results. Then a Clip Contrastive loss function is further proposed to align the extracted tumor features along both the spatial-temporal dimensions to improve the segmentation accuracy. In addition, the Global Retentive Memory is proposed to maintain the complementary tumor features with lower computing resources which can generate coherent temporal movement features. In this way, our model can significantly improve the spatial-temporal perception ability without increasing a large number of parameters, achieving more accurate segmentation results while maintaining a faster segmentation speed. Finally, we conduct extensive experiments to evaluate our proposed model on several video object segmentation datasets, the results show that our framework outperforms state-of-theart segmentation methods.

Using multimodal ultrasound including full-time-series contrast-enhanced ultrasound cines for identifying the nature of thyroid nodules.

Based on the conventional ultrasound images of thyroid nodules, contrast-enhanced ultrasound (CEUS) videos were analyzed to investigate whether CEUS improves the classification accuracy of benign and malignant thyroid nodules using machine learning (ML) radiomics and compared with radiologists. The B-mode ultrasound (B-US), real-time elastography (RTE), color doppler flow images (CDFI) and CEUS cines of patients from two centers were retrospectively gathered. Then, the region of interest (ROI) was delineated to extract the radiomics features. Seven ML algorithms combined with four kinds of radiomics data (B-US, B-US + CDFI + RTE, CEUS, and B-US + CDFI + RTE + CEUS) were applied to establish 28 models. The diagnostic performance of ML models was compared with interpretations from expert and nonexpert readers. A total of 181 thyroid nodules from 181 patients of 64 men (mean age, 42 years +/- 12) and 117 women (mean age, 46 years +/- 12) were included. Adaptive boosting (AdaBoost) achieved the highest area under the receiver operating characteristic curve (AUC) of 0.89 in the test set among 28 models when combined with B-US + CDFI + RTE + CEUS data and an AUC of 0.72 and 0.66 when combined with B-US and B-US + CDFI + RTE data. The AUC achieved by senior and junior radiologists was 0.78 versus (vs.) 0.69 (p > 0.05), 0.79 vs. 0.64 (p < 0.05), and 0.88 vs. 0.69 (p < 0.05) combined with B-US, B-US+CDFI+RTE and B-US+CDFI+RTE+CEUS, respectively. With the addition of CEUS, the diagnostic performance was enhanced for all seven classifiers and senior radiologists based on conventional ultrasound images, while no enhancement was observed for junior radiologists. The diagnostic performance of ML models was similar to senior radiologists, but superior to those junior radiologists.

Metric-Based Frame Selection and Deep Learning Model With Multi-Head Self Attention for Classification of Ultrasound Lung Video Images

Metric-Based Frame Selection and Deep Learning Model With Multi-Head Self Attention for Classification of Ultrasound Lung Video Images

Joint lesion detection and classification of breast ultrasound video via a clinical knowledge-aware framework

Joint lesion detection and classification of breast ultrasound video via a clinical knowledge-aware framework

Relationship Between Sensory Evaluations of Jelly Beverages and Neck Ultrasound Videos

Relationship Between Sensory Evaluations of Jelly Beverages and Neck Ultrasound Videos

Deformation analysis of forearm soft tissues surrounding the cephalic vein under shear on the skin surface using ultrasound video images

Deformation analysis of forearm soft tissues surrounding the cephalic vein under shear on the skin surface using ultrasound video images

Fetal growth analysis from ultrasound videos based on different biometrics using optimal segmentation and hybrid classifier.

Birth defects and their associated deaths, high health and financial costs of maternal care and associated morbidity are major contributors to infant mortality. If permitted by law, prenatal diagnosis allows for intrauterine care, more complicated hospital deliveries, and termination of pregnancy. During pregnancy, a set of measurements is commonly used to monitor the fetal health, including fetal head circumference, crown-rump length, abdominal circumference, and femur length. Because of the intricate interactions between the biological tissues and the US waves mother and fetus, analyzing fetal US images from a specialized perspective is difficult. Artifacts include acoustic shadows, speckle noise, motion blur, and missing borders. The fetus moves quickly, body structures close, and the weeks of pregnancy vary greatly. In this work, we propose a fetal growth analysis through US image of head circumference biometry using optimal segmentation and hybrid classifier. First, we introduce a hybrid whale with oppositional fruit fly optimization (WOFF) algorithm for optimal segmentation of segment fetal head which improves the detection accuracy. Next, an improved U-Net design is utilized for the hidden feature (head circumference biometry) extraction which extracts features from the segmented extraction. Then, we design a modified Boosting arithmetic optimization (MBAO) algorithm for feature optimization to selects optimal best features among multiple features for the reduction of data dimensionality issues. Furthermore, a hybrid deep learning technique called bi-directional LSTM with convolutional neural network (B-LSTM-CNN) for fetal growth analysis to compute the fetus growth and health. Finally, we validate our proposed method through the open benchmark datasets are HC18 (Ultrasound image) and oxford university research archive (ORA-data) (Ultrasound video frames). We compared the simulation results of our proposed algorithm with the existing state-of-art techniques in terms of various metrics.