Cine Clips Research Articles

This is another ultrasound cine clip demonstrating a fine-needle aspiration of an axillary lymph node without (left) and with (right) steering.

This is another ultrasound cine clip demonstrating a fine-needle aspiration of an axillary lymph node without (left) and with (right) steering.

This ultrasound cine clip demonstrates a fine-needle aspiration of an axillary lymph node without (left) and with (right) steering.

This ultrasound cine clip demonstrates a fine-needle aspiration of an axillary lymph node without (left) and with (right) steering.

Human Observer Sensitivity to Temporal Noise During B-Mode Ultrasound Scanning: Characterization and Imaging Implications.

This work measures temporal signal-to-noise ratio (SNR) thresholds that indicate when random noise during ultrasound scanning becomes imperceptible to expert human observers. Visible noise compromises image quality and can potentially lead to non-diagnostic scans. Noise can arise from both stable acoustic sources (clutter) or randomly varying electronic sources (temporal noise). Extensive engineering effort has focused on decreasing noise in both of these categories. In this work, an observer study with five practicing sonographers was performed to assess sonographer sensitivity to temporal noise in ultrasound cine clips. Understanding the conditions where temporal noise is no longer visible during ultrasound imaging can inform engineering efforts seeking to minimize the impact this noise has on image quality. The sonographers were presented with paired temporal noise-free and noise-added simulated speckle cine clips and asked to select the noise-added clips. The degree of motion in the imaging target was found to have a significant effect on the SNR levels where noise was perceived, while changing imaging frequency had little impact. At realistic in vivo motion levels, temporal noise was not perceived in cine clips at and above 28 dB SNR. In a case study presented here, the potential of adaptive intensity adjustment based on this noise perception threshold is validated in a fetal imaging scenario. This study demonstrates how noise perception thresholds can be applied to help design or tune ultrasound systems for different imaging tasks and noise conditions.

Cine clips increase the detection of thyroid pyramidal lobe in routine thyroid sonogram.

Identification and resection of the thyroid pyramidal lobe is important for thyroid cancer surgery in order to prevent interval cancer in residual thyroid tissue. The purpose of this study was to determine how often a thyroid pyramidal lobe is found in patients with and without previous thyroidectomy and to optimise the protocol for identifying thyroid pyramidal lobes during routine thyroid ultrasonography. In this prospective study, a total of 1579 patients who received routine thyroid ultrasound scans at a single centre were enrolled. A dedicated standard scanning protocol was established containing both static images of the anterior neck superior to the thyroid as well as a transverse cine loop starting from the isthmus to the hyoid bone. The presence and features of thyroid pyramidal lobes were evaluated and compared. Detection rate of thyroid pyramidal lobes in patients without thyroidectomy improved from 39.5% (480/1215) to 49.7% (640/1215) with protocol adding cine-loop as compared to protocol without cine-loop. The cine-loops were particularly helpful in the detection of thyroid pyramidal lobes when it is separated from the main lobe or in thyroidectomy patients. By adding assessment with cine-loop into the dedicated protocol, we have further detected different pathologies occurring on thyroid pyramidal lobes including benign and malignant solid nodules and pseudo-nodules of Hashimoto's thyroiditis. The addition of dynamic assessment with cine-loop increases the detection rate of thyroid pyramidal lobes. By paying attention to the thyroid pyramidal lobe in pre-operative diagnostic sonographic images, we can help to avoid incomplete removal of the thyroid gland during thyroidectomy.

Transfer Learning-Based B-Line Assessment of Lung Ultrasound for Acute Heart Failure

ObjectiveB-lines assessed by lung ultrasound (LUS) outperform physical exam, chest radiograph, and biomarkers for the associated diagnosis of acute heart failure (AHF) in the emergent setting. The use of LUS is however limited to trained professionals and suffers from interpretation variability. The objective was to utilize transfer learning to create an AI-enabled software that can aid novice users to automate LUS B-line interpretation. MethodsData from an observational AHF LUS study provided standardized cine clips for AI model development and evaluation. A total of 49,952 LUS frames from 30 patients were hand scored and trained on a convolutional neural network (CNN) to interpret B-lines at the frame level. A random independent evaluation set of 476 LUS clips from 60 unique patients assessed model performance. The AI models scored the clips on both a binary and ordinal 0–4 multiclass assessment. ResultsA multiclassification AI algorithm had the best performance at the binary level when applied to the independent evaluation set, AUC of 0.967 (95% CI 0.965–0.970) for detecting pathologic conditions. When compared to expert blinded reviewer, the 0–4 multiclassification AI algorithm scale had a reported linear weighted kappa of 0.839 (95% CI 0.804–0.871). ConclusionsThe multiclassification AI algorithm is a robust and well performing model at both binary and ordinal multiclass B-line evaluation. This algorithm has the potential to be integrated into clinical workflows to assist users with quantitative and objective B-line assessment for evaluation of AHF.

Artificial Intelligence for Breast Ultrasound: AJR Expert Panel Narrative Review.

Breast ultrasound is used in a wide variety of clinical scenarios, including both diagnostic and screening applications. Limitations of ultrasound, however, include its low specificity and, for automated breast ultrasound screening, the time necessary to review whole-breast ultrasound images. As of this writing, four AI tools that are approved or cleared by the FDA address these limitations. Current tools, which are intended to provide decision support for lesion classification and/or detection, have been shown to increase specificity among non-specialists and to decrease interpretation times. Potential future applications include triage of patients with palpable masses in low-resource settings, preoperative prediction of axillary lymph node metastasis, and preoperative prediction of neoadjuvant chemotherapy response. Challenges in the development and clinical deployment of AI for ultrasound include: the limited availability of curated training datasets compared to mammography; the high variability in ultrasound image acquisition due to equipment- and operator-related factors (which may limit algorithm generalizability); and the lack of post-implementation evaluation studies. Furthermore, current AI tools for lesion classification were developed based on 2D data, but diagnostic accuracy could potentially be improved if multimodal ultrasound data were used, such as color Doppler, elastography, cine clips, and 3D imaging.

Is myocardial fibrosis appropriately assessed by 2D strain derived integrated backscatter?

Abstract Introduction Calibrated integrated backscatter (cIB) may be obtained from bidimensional (2D) strain echocardiography as a quantification measurement of myocardial ultrasound reflectivity. Increased collagen content of the myocardium modifies tissue reflectivity and cIB is suggested as a marker of left ventricular (LV) fibrosis. However, its diagnostic accuracy is not fully established. Aim To assess the correlation between cIB by 2D strain and LV myocardial fibrosis (MF), as evaluated by non-invasive imaging and histopathology. Methods Correlation study from a cohort of 157 patients with severe symptomatic aortic stenosis (AS) referred for surgical aortic valve replacement (AVR). Patients with complete preoperative transthoracic echocardiography, cardiac magnetic resonance (CMR) and endomyocardial biopsy (EMB) obtained from the anterior basal septum at the time of AVR were selected. Two groups of 30 patients were evaluated, with and without late gadolinium enhancement (LGE) at CMR. Integrated Backscatter was obtained at QRS peak in decibels (dB) from both parasternal long axis (PLAX) and apical-three-chamber (A3C) cine clips at Qanalysis. cIB was calculated by subtracting the pericardial intensity from the average of the anteroseptal and basal inferolateral wall values. Correlation analysis was performed for the whole group of patients with global and segmental (anterior basal septum) values of native T1 and extracellular volume (ECV), and EMB collagen volume fraction (CVF) from Masson´s Trichrome staining. IBS values were compared in both group of patients. Results 60 patients (74 [36-74] years, 45% male) with high gradient (mean gradient: 63±20mmHg), normal flow (45±10mL/m2), preserved ejection fraction (60±9%) AS were included. Basal septal cIB was –9.17±9.45dB and –17.45 (–31.2–10.95) dB from A3C and PLAX views, respectively. These indexes did not correlate with basal septum thickness or global LV mass. Absolute and cIB did not correlate neither with global and regional T1 and ECV values, nor with CVF at EMB. These were not significantly different in both groups of patients and there was no correlation between cIB values and mass of replacement MF in patients with LGE. Conclusion In these cohort of patients with classical severe AS, there was no correlation of cIB with imaging markers of both replacement and diffuse MF. cIB also didn´t correlate with CVF at histopathology. These findings suggest that reflectivity indexes are not suitable for myocardial tissue characterization in this setting.

Variability in Interpretation of Echocardiographic Signs of Tamponade: A Survey of Emergency Physician Sonographers

BackgroundCardiac tamponade is associated with high mortality, and making the diagnosis is a core skill of emergency physicians. Proper diagnosis relies on specific clinical and echocardiographic findings. It is not known whether expert sonographers consistently recognize echocardiographic signs of tamponade. ObjectivesTo assess whether expert sonographers agree on echocardiographic signs of tamponade. MethodsA 20-question survey consisting of 18 cine loops and 2 still images was distributed to the Academy of Emergency Ultrasound Section of the Society for Academic Emergency Medicine. Respondents answered “yes” or “no” to whether there was echocardiographic evidence of tamponade. Subgroup analyses of demographics and echocardiographic views were reported. The data were analyzed using Krippendorff's alpha (α) to assess interrater reliability (IRR) between respondents. ResultsEighty-four physicians responded and 56 completed the survey. All partial and completed surveys were analyzed. The overall IRR was poor (α = 0.60, 95% confidence interval [CI] 0.44–0.76). Residency graduation within 5 years (α = 0.66, 95% CI 0.5–0.8) was associated with higher IRR compared with those > 5 years (α = 0.53, 95% CI 0.37–0.69). The highest IRR was observed when images of mitral valve inflow pulse-wave Doppler (α = 0.81, CI 0.70–0.92) were used and the poorest IRR was on images from the parasternal short view (α = 0.28, 95% CI 0.05–0.49). ConclusionThere was poor agreement among expert emergency medicine sonographers in identifying echocardiographic signs of cardiac tamponade from a single cine loop or clip without clinical context. Further investigation is warranted to understand differences in recognition of clinical tamponade.

Thyroid Cancer Polygenic Risk Score Improves Classification of Thyroid Nodules as Benign or Malignant.

Thyroid nodule ultrasound-based risk stratification schemas rely on the presence of high-risk sonographic features. However, some malignant thyroid nodules have benign appearance on thyroid ultrasound. New methods for thyroid nodule risk assessment are needed. We investigated polygenic risk score (PRS) accounting for inherited thyroid cancer risk combined with ultrasound-based analysis for improved thyroid nodule risk assessment. The convolutional neural network classifier was trained on thyroid ultrasound still images and cine clips from 621 thyroid nodules. Phenome-wide association study (PheWAS) and PRS PheWAS were used to optimize PRS for distinguishing benign and malignant nodules. PRS was evaluated in 73 346 participants in the Colorado Center for Personalized Medicine Biobank. When the deep learning model output was combined with thyroid cancer PRS and genetic ancestry estimates, the area under the receiver operating characteristic curve (AUROC) of the benign vs malignant thyroid nodule classifier increased from 0.83 to 0.89 (DeLong, P value = .007). The combined deep learning and genetic classifier achieved a clinically relevant sensitivity of 0.95, 95% CI [0.88-0.99], specificity of 0.63 [0.55-0.70], and positive and negative predictive values of 0.47 [0.41-0.58] and 0.97 [0.92-0.99], respectively. AUROC improvement was consistent in European ancestry-stratified analysis (0.83 and 0.87 for deep learning and deep learning combined with PRS classifiers, respectively). Elevated PRS was associated with a greater risk of thyroid cancer structural disease recurrence (ordinal logistic regression, P value = .002). Augmenting ultrasound-based risk assessment with PRS improves diagnostic accuracy.

Is myocardial fibrosis appropriately assessed by 2D strain derived integrated backscatter?

Abstract Funding Acknowledgements Type of funding sources: None. Introduction Calibrated integrated backscatter (cIBS) may be obtained from bidimensional (2D) strain echocardiography as a quantification measurement of myocardial ultrasound reflectivity. Increased collagen content of the myocardium modifies tissue reflectivity and cIBS is suggested as a marker of left ventricular (LV) fibrosis. However, its diagnostic accuracy is not established. Aim To assess the correlation between cIBS by 2D strain and LV myocardial fibrosis (MF), as evaluated by non-invasive imaging and histopathology. Methods Correlation study from a cohort of 157 patients with severe symptomatic aortic stenosis (AS) referred for surgical aortic valve replacement (AVR). Patients with complete preoperative transthoracic echocardiography, cardiac magnetic resonance (CMR) and endomyocardial biopsy (EMB) obtained from the anterior basal septum at the time of AVR were selected. Two groups of 30 patients were evaluated, with and without late gadolinium enhancement (LGE) at CMR. IBS was obtained at QRS peak in decibels (dB) from both parasternal long axis (PLAX) and apical three chamber (A3C) cine clips at Qanalysis (Figure 1A). cIBS was calculated by subtracting the pericardial intensity from the average of the anteroseptal and basal inferolateral wall values. Correlation analysis was performed for the whole group of patients with global and segmental (anterior basal septum) values of native T1 and extracellular volume (ECV), and EMB collagen volume fraction (CVF) from Masson´s Trichrome staining. IBS values were compared in both group of patients. Results 60 patients (73 [68–74] years, 45% male) with high gradient (mean gradient: 64±20mmHg), normal flow (45±10mL/m2), preserved ejection fraction (60±9%) AS. Basal septal cIBS was −9.2±9.5dB and −16.3±7.9dB from A3C and PLAX views, respectively. These indexes did not correlate with basal septum thickness or global LV mass. Absolute and cIBS did not correlate neither with global and regional T1 and ECV values, nor with CVF at EMB (Figure 1B). These were not significantly different in both groups of patients and there was no correlation between cIBS values and mass of replacement MF in patients with LGE. Conclusion In these cohort of patients with classical severe AS, there was no correlation of cIBS with imaging markers of both replacement and diffuse MF. cIBS also didn´t correlate with CVF at histopathology. These findings suggest that reflectivity indexes are not suitable for myocardial tissue characterization in this setting.

Development and interobserver reliability of a rating scale for lung ultrasound pathology in lower respiratory tract infection

ObjectivesThe purpose of this study was to develop a severity rating scale for lung ultrasound pathology in lower respiratory tract infection based on multicenter expert consensus, and to test inter-rater reliability. MethodsTen point-of-care ultrasound experts from three academic institutions developed the scale iteratively through literature review, expert opinion, and pilot testing. Clips were prospectively collected from adults suspected of COVID-19 using a 14-zone scanning protocol. Blinded reviewers independently rated four data subsets. The rating scale was refined through eight consensus-building discussions reviewing challenging cases from the first three subsets. The final scale consisted of a set of ordinal scores from 0 to 4, for five sonographic findings: B-lines, pleural line abnormalities, consolidations, pleural effusions, and overall lung aeration. Ratings from the fourth subset were analyzed to determine reliability based on intraclass correlation coefficient (ICC). A total of 11,126 cine clips from 220 patients were acquired. DiscussionAfter excluding uninterpretable clips, the test dataset contained 81 clips and yielded an average ICC of 0.70 across the five sonographic findings (0.76 for B-lines, 0.52 for pleural line abnormalities, 0.71 for consolidations, 0.80 for pleural effusions, and 0.70 for overall lung aeration). Improvements in agreement were observed with each successive review session and dataset rating. ConclusionThe lung ultrasound severity scale established by multicenter expert consensus achieved moderate to good inter-rater reliability. The scale could be used clinically to standardize assessment of lower respiratory tract infection and in future studies to develop methods for automated interpretation of lung ultrasound pathology.

Contrast-enhanced subharmonic aided pressure estimation for assessment of intracranial pressure in vivo.

Intracranial pressure (ICP) monitoring in children currently requires invasive techniques. Subharmonic aided pressure estimation (SHAPE) uses contrast-enhanced ultrasound (CEUS) to measure intravascular and interstitial pressure, but utility in ICP measurements has yet to be explored. The objective of this study was to investigate SHAPE as a novel tool for noninvasive ICP measurements in fetal lambs. Eighteen fetal lambs at 107-139days gestational age (term = 145days) underwent subdural ICP catheter placement. The brain was imaged in the coronal plane in CEUS mode optimized for SHAPE, while infusing an US contrast agent into the fetal circulation. After SHAPE calibration, saline was infused via the subdural catheter to increase ICP. Five-second SHAPE cine clips were obtained at various ICPs. Subharmonic intensity values of the whole brain and thalami were correlated with ICP values using mixed effects linear regression analyses and the strength of the relationship was evaluated by Spearman's rank-order correlation. Forty-nine experiments produced 723 datapoints, including SHAPE intensity values and mean ICP measurements. There was a statistically significant inverse relationship between SHAPE intensity values and ICP measurements in the whole brain and thalami (median rho value - 0.58 and - 0.56, respectively). SHAPE intensity values of the brain demonstrate an inverse and statistically significant correlation with in vivo ICP measurements in an animal model.

Diagnosis of Pregnancy Complications Using Blind Ultrasound Sweeps Performed by Individuals Without Prior Formal Ultrasound Training.

To estimate the diagnostic accuracy of blind ultrasound sweeps performed with a low-cost, portable ultrasound system by individuals with no prior formal ultrasound training to diagnose common pregnancy complications. This is a single-center, prospective cohort study conducted from October 2020 to January 2022 among people with second- and third-trimester pregnancies. Nonspecialists with no prior formal ultrasound training underwent a brief training on a simple eight-step approach to performing a limited obstetric ultrasound examination that uses blind sweeps of a portable ultrasound probe based on external body landmarks. The sweeps were interpreted by five blinded maternal-fetal medicine subspecialists. Sensitivity, specificity, and positive and negative predictive values for blinded ultrasound sweep identification of pregnancy complications (fetal malpresentation, multiple gestations, placenta previa, and abnormal amniotic fluid volume) were compared with a reference standard ultrasonogram as the primary analysis. Kappa for agreement was also assessed. Trainees performed 194 blinded ultrasound examinations on 168 unique pregnant people (248 fetuses) at a mean of 28±5.85 weeks of gestation for a total of 1,552 blinded sweep cine clips. There were 49 ultrasonograms with normal results (control group) and 145 ultrasonograms with abnormal results with known pregnancy complications. In this cohort, the sensitivity for detecting a prespecified pregnancy complication was 91.7% (95% CI 87.2-96.2%) overall, with the highest detection rate for multiple gestations (100%, 95% CI 100-100%) and noncephalic presentation (91.8%, 95% CI 86.4-97.3%). There was high negative predictive value for placenta previa (96.1%, 95% CI 93.5-98.8%) and abnormal amniotic fluid volume (89.5%, 95% CI 85.3-93.6%). There was also substantial to perfect mean agreement for these same outcomes (range 87-99.6% agreement, Cohen κ range 0.59-0.91, P<.001 for all). Blind ultrasound sweeps of the gravid abdomen guided by an eight-step protocol using only external anatomic landmarks and performed by previously untrained operators with a low-cost, portable, battery-powered device had excellent sensitivity and specificity for high-risk pregnancy complications such as malpresentation, placenta previa, multiple gestations, and abnormal amniotic fluid volume, similar to results of a diagnostic ultrasound examination using a trained ultrasonographer and standard-of-care ultrasound machine. This approach has the potential to improve access to obstetric ultrasonography globally.

Evaluation of the Cerebrospinal Fluid Flow Dynamics with Microvascular Imaging Ultrasound in Infants.

Microvascular imaging ultrasound (MVI) can detect slow blood flow in small-caliber cerebral vessels. This technology may help assess flow in other intracranial structures, such as the ventricular system. In this study, we describe the use of MVI for characterizing intraventricular cerebrospinal fluid (CSF) flow dynamics in infants. We included infants with brain ultrasound that had MVI B-Flow cine clips in the sagittal plane. Two blinded reviewers examined the images, dictated a diagnostic impression, and identified the third ventricle, cerebral aqueduct, fourth ventricle, and CSF flow direction. A third reviewer evaluated the discrepancies. We evaluated the association of visualization of CSF flow as detectable with MVI, with the diagnostic impressions. We also assessed the inter-rater reliability (IRR) for detecting CSF flow. We evaluated 101 infants, mean age 40 ± 53 days. Based on brain MVI B-Flow, a total of 49 patients had normal brain US scans, 40 had hydrocephalus, 26 had intraventricular hemorrhage (IVH), and 14 had hydrocephalus+IVH. Using spatially moving MVI signal in the third ventricle, cerebral aqueduct, and fourth ventricle as the criteria for CSF flow, CSF flow was identified in 10.9% (n = 11), 15.8% (n = 16), and 16.8% (n = 17) of cases, respectively. Flow direction was detected in 19.8% (n = 20) of cases; 70% (n = 14) was caudocranial, 15% (n = 3) was craniocaudal, and 15% (n = 3) bidirectional, with IRR = 0.662, p < 0.001. Visualization of CSF flow was significantly associated with the presence of IVH alone (OR 9.7 [3.3-29.0], p < 0.001) and IVH+hydrocephalus (OR 12.4 [3.5-440], p < 0.001), but not with hydrocephalus alone (p = 0.116). This study demonstrates that MVI can detect CSF flow dynamics in infants with a history of post-hemorrhagic hydrocephalus with a high IRR.

Abstract 12868: Real-Time Explainable Detection of Stenosis in Coronary Angiography

Background: Coronary angiography is the gold standard for evaluating coronary artery disease and determining the appropriate treatment strategy for a patient. The aim of this study is to use a machine learning (ML)-based object detection algorithm to identify images with clinically significant stenoses and provide explainable output by localizing the area of stenosis within each image. Methods: We utilize the YOLOv5 object detection algorithm to detect clinically significant stenoses by treating narrowing of the vessel as the object to be located in an image. Models are trained utilizing a 5-fold cross-validation strategy on a dataset consisting of the middle frame from 756 randomly selected cine clips from an interventional cardiology suite. Maximum object confidence on the validation dataset is used for downstream whole image classification. Results: Across 5 folds, mean bounding box precision was 0.342 ± 0.074 and recall was 0.305 ± 0.021 at an intersection over union threshold of 0.6. The image classification model achieved a mean and standard deviation area under the receiver operating characteristic curve of 0.863 ± 0.042 with a mean accuracy of 0.809 ± 0.041 at the optimal confidence cutoff of 0.283 ± 0.241. Additionally, trained model predictions occurred at a rate of 32.8 ± 3.6 frames per second, allowing for live model inference and provider support at a standard angiographic frame rate of 15 frames per second. Conclusions: The proposed object detection ML approach is a time-efficient and data-efficient method of diagnosing significant coronary artery stenosis. These results combine state-of-the-art performance in detecting stenosis in angiographic images with improved explainability over single-step classification with a deep neural network.

RF11 | PSAT234 Deep learning analysis of thyroid nodule ultrasound images has high sensitivity and negative predictive value to rule-out thyroid cancer

Abstract Purpose To evaluate deep learning analysis of thyroid nodule ultrasound images as a rule-out test for thyroid malignancy. Methods Supervised deep learning (DL) classifier of thyroid nodules was trained on 32,545 thyroid US images from 621 nodules representing all major benign and malignant types of thyroid lesions and tested on an independent set of 145 nodules collected at a different healthcare system in the United States. The Big Transfer BiT-M ResNet-50×1 convolutional neural net architecture was modified to contain 3, 4, 6 and 3 PreActBottleneck units per block 1 through 4. Weights pretrained on the ImageNet-21k dataset were loaded and weights for blocks 3 and 4 were fine-tuned for the binary classification task of distinguishing benign and malignant thyroid nodules. Results The deep learning thyroid nodule classifier achieved an area under receiver operating characteristic curve (AUROC) of 0.889 on five-fold cross-validation. The AUROC improved when images were scaled by nodule size and six randomly selected cine clip frames were added to the training set per epoch. GradCAM class activation heatmaps revealed that microcalcifications and spongiform appearance were reliably recognized by the classifier as malignant and benign features, respectively. Spongiform nodules were found to be benign even when microcystic spaces constituted less than 50% of nodule volume. To investigate the clinical relevance of the benign vs. malignant classifier, the binary classification threshold for the probability of malignancy generated by model was set at 7% to achieve sensitivity and negative predictive value (NPV) comparable to that of the fine needle aspiration biopsy (FNA). At this threshold, cross-validated deep-learning model achieved a sensitivity of 90%, specificity of 63%, positive predictive value (PPV) of 46% and negative predictive value of 94%. When tested on an independent image set that includes 18 classic papillary thyroid cancers (PTC), 5 follicular variant PTC, 4 medullary thyroid cancers, 3 follicular thyroid cancers (FTC), and 1 Hurthle cell thyroid cancer, the DL classifier achieved AUROC of 0.88, sensitivity of 97%, specificity of 61%, PPV of 40% and NPV of 99%. A single minimally-invasive FTC that had no suspicious features on thyroid ultrasound was incorrectly classified as benign. Conclusions This study demonstrates that the ultrasound-based deep-learning classifier of thyroid nodules achieves sensitivity and negative predictive value comparable to that of thyroid fine needle aspiration (FNA). Clinicians may use this tool to augment clinical judgment when determining whether to perform FNA procedures. Presentation: Saturday, June 11, 2022 1:00 p.m. - 3:00 p.m., Saturday, June 11, 2022 1:06 p.m. - 1:11 p.m.

161 Multicenter Interobserver Agreement of Lung Ultrasound Findings in COVID-19 Patients

161 Multicenter Interobserver Agreement of Lung Ultrasound Findings in COVID-19 Patients

Panoramic Reconstruction of B-mode Lung Ultrasound Images Acquired using a Longitudinal Volume Sweep Imaging Protocol.

The ongoing COVID-19 pandemic has already affected more than 300 million people worldwide. Medical imaging shortage affects an estimated of 4 billion people, especially in rural and remote areas (RAs), limiting diagnostic assessment of respiratory illness. Lung ultrasound imaging (LUS) together with volume sweep imaging (VSI) acquisition protocols have been successfully piloted as a solution for lung screening in RAs eliminating the need for trained operators and on-site radiologists. Nevertheless, this protocol requires the acquisition of 12 videos for 6 areas with both longitudinal and transverse positions of the transducer. Nonetheless, bandwidth limitations can hamper the transmission of these videos for remote interpretation. This work aimed to developed a stitching algorithm capable of generating a panoramic reconstruction of LUS cine clips. The results show reconstructions with minimal loss of information as 92.5% of the panoramic images conserved the presence of A-lines. These results show that LUS can be represented as an image without significantly compromising its quality. This can be useful to overcome bandwidth issues as well as improve the time on lung assessment of the patient.

Cardiac Size, Shape, and Ventricular Contractility in Fetuses at Sea Level With an Estimated Weight Less‐than 10th Centile

To investigate cardiac size, shape, and ventricular contractility in fetuses with estimated fetal weight (EFW) <10th centile at sea level (Houston). A prospective ultrasound study examined 37 fetuses with EFW <10th centile at sea level. High-frequency cine clips were used to evaluate the 4-chamber view including end-diastolic measurements and global sphericity index. The size, shape, and contractility of both ventricles were analyzed with speckle tracking methods. Z scores were calculated using the mean ±standard deviation (SD) derived from normal controls. Measurements were abnormal if their Z score values were <-1.65 or >+1.65. The proportion of small fetuses with abnormal parameters was compared to normal reference ranges. Results were compared to a similar published study of small fetuses at higher altitude in Denver. About one-third of Houston fetuses with EFW <10th centile had enlarged globular shaped 4-chamber hearts with increased right ventricle (RV) area, RV basal-apical length, RV base width, and left ventricle (LV) basal-apical length measurements. Bilateral ventricular hypertrophy was often present. An increased proportion of Houston fetuses had increased ventricular contractility. However, decreased ventricular contractility was more prevalent for higher altitude fetuses. Third trimester fetuses at sea level, with an EFW <10th centile, were often associated with enlarged and globular-shaped hearts. They had increased global and longitudinal ventricular contractility as compared to controls. Higher altitude fetuses also had enlarged globular-shaped hearts but with a greater proportion of cases having decreased ventricular contractility as compared to the sea level cohort.

Characterizing breast masses using an integrative framework of machine learning and CEUS-based radiomics.

We evaluated the performance of contrast-enhanced ultrasound (CEUS) based on radiomics analysis to distinguish benign from malignant breast masses. 131 women with suspicious breast masses (BI-RADS 4a, 4b, or 4c) who underwent CEUS examinations (using intravenous injection of perflutren lipid microsphere or sulfur hexafluoride lipid-type A microspheres) prior to ultrasound-guided biopsies were retrospectively identified. Post biopsy pathology showed 115 benign and 16 malignant masses. From the cine clip of the CEUS exams obtained using the built-in GE scanner software, breast masses and adjacent normal tissue were then manually segmented using the ImageJ software. One frame representing each of the four phases: precontrast, early, peak, and delay enhancement were selected post segmentation from each CEUS clip. 112 radiomic metrics were extracted from each segmented tissue normalized breast mass using custom Matlab® code. Linear and nonlinear machine learning (ML) methods were used to build the prediction model to distinguish benign from malignant masses. tenfold cross-validation evaluated model performance. Area under the curve (AUC) was used to quantify prediction accuracy. Univariate analysis found 35 (38.5%) radiomic variables with p < 0.05 in differentiating between benign from malignant masses. No feature selection was performed. Predictive models based on AdaBoost reported an AUC = 0.72 95% CI (0.56, 0.89), followed by Random Forest with an AUC = 0.71 95% CI (0.56, 0.87). CEUS based texture metrics can distinguish between benign and malignant breast masses, which can, in turn, lead to reduced unnecessary breast biopsies.