Spatiotemporal Deep Learning-Based Cine Loop Quality Filter for Handheld Point-of-Care Echocardiography.

The reliability of automated image interpretation of point-of-care (POC) echocardiography scans depends on the quality of the acquired ultrasound data. This work reports on the development and validation of spatiotemporal deep learning models to assess the suitability of input ultrasound cine loops collected using a handheld echocardiography device for processing by an automated quantification algorithm (e.g. ejection fraction estimation). POC echocardiograms (n=885 DICOM cine loops from 175 patients) from two sites were collected using a handheld ultrasound device and annotated for image quality at the frame-level. Attributes of high-quality frames for left ventricular (LV) quantification included a temporally-stable LV, reasonable coverage of LV borders, and good contrast between the borders and chamber. Attributes of low-quality frames included temporal instability of the LV and/or imaging artifacts (e.g., lack of contrast, haze, reverberation, acoustic shadowing). Three different neural network architectures were investigated - (a) frame-level convolutional neural network (CNN) which operates on individual echo frames (VectorCNN), (b) single-stream sequence-level CNN which operates on a sequence of echo frames (VectorCNN+LSTM) and (c) two-stream sequence-level CNNs which operate on a sequence of echo and optical flow frames (VectorCNN+LSTM+Average, VectorCNN+LSTM+MinMax, and VectorCNN+LSTM+ConvPool). Evaluation on a sequestered test dataset containing 76 DICOM cine loops with 16,914 frames showed that VectorCNN+LSTM can effectively utilize both spatial and temporal information to regress the quality of an input frame (accuracy: 0.925, sensitivity = 0.860, specificity = 0.952), compared to the frame-level VectorCNN that only utilizes spatial information in that frame (accuracy: 0.903, sensitivity = 0.791, specificity = 0.949). Furthermore, an independent sample t-test indicated that the cine loops classified to be of adequate quality by the VectorCNN+LSTM model had a statistically significant lower bias in the automatically estimated EF (mean bias = - 3.73 ± 7.46 %, versus a clinically obtained reference EF) compared to the loops classified as inadequate (mean bias = -15.92 ± 12.17 %) (p = 0.007). Thus, cine loop stratification using the proposed spatiotemporal CNN model improves the reliability of automated point-of-care echocardiography image interpretation.

Artificial intelligence-guided detection of under-recognized cardiomyopathies on point-of-care cardiac ultrasound: a multi-center study.

BACKGROUND: Point-of-care ultrasonography (POCUS) enables cardiac imaging at the bedside and in communities but is limited by abbreviated protocols and variation in quality. We developed and tested artificial intelligence (AI) models to automate the detection of under-diagnosed cardiomyopathies from cardiac POCUS. METHODS: In a development set of 290,245 transthoracic echocardiographic videos across the Yale-New Haven Health System (YNHHS), we used augmentation approaches and a customized loss function weighted for view quality to derive a POCUS-adapted, multi-label, video-based convolutional neural network (CNN) that discriminates HCM (hypertrophic cardiomyopathy) and ATTR-CM (transthyretin amyloid cardiomyopathy) from controls without known disease. We evaluated the final model across independent, internal and external, retrospective cohorts of individuals who underwent cardiac POCUS across YNHHS and Mount Sinai Health System (MSHS) emergency departments (EDs) (2011-2024) to prioritize key views and validate the diagnostic and prognostic performance of single-view screening protocols. FINDINGS: We identified 33,127 patients (median age 61 [IQR: 45-75] years, n=17,276 [52.2%] female) at YNHHS and 5,624 (57 [IQR: 39-71] years, n=1,953 [34.7%] female) at MSHS with 78,054 and 13,796 eligible cardiac POCUS videos, respectively. An AI-enabled single-view screening approach successfully discriminated HCM (AUROC of 0.90 [YNHHS] & 0.89 [MSHS]) and ATTR-CM (YNHHS: AUROC of 0.92 [YNHHS] & 0.99 [MSHS]). In YNHHS, 40 (58.0%) HCM and 23 (47.9%) ATTR-CM cases had a positive screen at median of 2.1 [IQR: 0.9-4.5] and 1.9 [IQR: 1.0-3.4] years before clinical diagnosis. Moreover, among 24,448 participants without known cardiomyopathy followed over 2.2 [IQR: 1.1-5.8] years, AI-POCUS probabilities in the highest (vs lowest) quintile for HCM and ATTR-CM conferred a 15% (adj.HR 1.15 [95%CI: 1.02-1.29]) and 39% (adj.HR 1.39 [95%CI: 1.22-1.59]) higher age- and sex-adjusted mortality risk, respectively. INTERPRETATION: We developed and validated an AI framework that enables scalable, opportunistic screening of treatable cardiomyopathies wherever POCUS is used.

Emergency Physician-performed Echocardiogram in Non-ST Elevation Acute Coronary Syndrome Patients Requiring Coronary Intervention.

Introduction: Identification of patients not meeting catheterization laboratory activation criteria by electrocardiogram (ECG) but who would benefit from early coronary intervention remains challenging in the emergency department (ED). The purpose of this study was to evaluate whether emergency physician (EP)-performed point-of-care transthoracic echocardiography (POC TTE) could help identify patients who required coronary intervention within this population. Methods: This was a retrospective observational cohort study of adult patients who presented to two EDs between 2018-2020. Patients were included if they received a POC TTE and underwent diagnostic coronary angiography within 72 hours of ED presentation. We excluded patients meeting catheterization laboratory activation criteria on initial ED ECG. Ultrasound studies were independently reviewed for presence of regional wall motion abnormalities (RWMA) by two blinded ultrasound fellowship-trained EPs. We then calculated test characteristics for coronary intervention. Results: Of the 221 patient encounters meeting inclusion criteria, 104 (47%) received coronary intervention or coronary artery bypass grafting (CABG) referral. Overall prevalence of RWMA on POC TTE was 35% (95% confidence interval [CI] 29-42%). Presence of RWMA had 38% (95% CI 29-49%) sensitivity and 68% (95% CI 58-76%) specificity for coronary intervention/CABG referral. Presence of "new" RWMA (presence on EP-performed POC TTE and prior normal echocardiogram) had 43% (95% CI 10-82%) sensitivity and 93% (95% CI 66-100%) specificity for coronary intervention/CABG referral. The EP-performed POC TTE interpretation of RWMA had 57% (95% CI 47-67%) sensitivity and 96% (95% CI 87-100%) specificity for presence of RWMA on subsequent cardiology echocardiogram during the same admission. Conclusion: Presence of RWMA on EP-performed POC TTE had limited sensitivity or specificity for coronary intervention or referral to CABG. The observed specificity appeared to trend higher in subjects with a prior echocardiogram demonstrating absence of RWMA, although a larger sample size will be required to confirm this finding. The EP-performed POC TTE RWMA had high specificity for presence of RWMA on subsequent cardiology echocardiogram. Further evaluation of the diagnostic performance of new RWMA on EP-performed POC TTE with a dedicated cohort is warranted.

Machine Learning Algorithm Detection of Confluent B-Lines.

OBJECTIVE: B-lines are a ring-down artifact of lung ultrasound that arise with increased alveolar water in conditions such as pulmonary edema and infectious pneumonitis. Confluent B-line presence may signify a different level of pathology compared with single B-lines. Existing algorithms aimed at B-line counting do not distinguish between single and confluent B-lines. The objective of this study was to test a machine learning algorithm for confluent B-line identification. METHODS: This study used a subset of 416 clips from 157 subjects, previously acquired in a prospective study enrolling adults with shortness of breath at two academic medical centers, using a hand-held tablet and a 14-zone protocol. After exclusions, random sampling generated a total of 416 clips (146 curvilinear, 150 sector and 120 linear) for review. A group of five experts in point-of-care ultrasound blindly evaluated the clips for presence/absence of confluent B-lines. Ground truth was defined as majority agreement among the experts and used for comparison with the algorithm. RESULTS: Confluent B-lines were present in 206 of 416 clips (49.5%). Sensitivity and specificity of confluent B-line detection by algorithm compared with expert determination were 83% (95% confidence interval [CI]: 0.77-0.88) and 92% (95% CI: 0.88-0.96). Sensitivity and specificity did not statistically differ between transducers. Agreement between algorithm and expert for confluent B-lines measured by unweighted κ was 0.75 (95% CI: 0.69-0.81) for the overall set. CONCLUSION: The confluent B-line detection algorithm had high sensitivity and specificity for detection of confluent B-lines in lung ultrasound point-of-care clips, compared with expert determination.

Two- Versus 8-Zone Lung Ultrasound in Heart Failure: Analysis of a Large Data Set Using a Deep Learning Algorithm.

OBJECTIVE: Scanning protocols for lung ultrasound often include 8 or more lung zones, which may limit real-world clinical use. We sought to compare a 2-zone, anterior-superior thoracic ultrasound protocol for B-line artifact detection with an 8-zone approach in patients with known or suspected heart failure using a deep learning (DL) algorithm. METHODS: Adult patients with suspected heart failure and B-lines on initial lung ultrasound were enrolled in a prospective observational study. Subjects received daily ultrasounds with a hand-held ultrasound system using an 8-zone protocol (right and left anterior/lateral and superior/inferior). A previously published deep learning algorithm that rates severity of B-lines on a 0-4 scale was adapted for use on hand-held ultrasound full video loops. Average severities for 8 and 2 zones were calculated utilizing DL ratings. Bland-Altman plot analyses were used to assess agreement and identify bias between 2- and 8-zone scores for both primary (all patients, 5728 videos, 205 subjects) and subgroup (confirmed diagnosis of heart failure or pulmonary edema, 4464 videos, 147 subjects) analyses. RESULTS: Bland-Altman plot analyses revealed excellent agreement for both primary and subgroup analyses. The absolute difference on the 4-point scale between 8- and 2-zone average scores was not significant for the primary dataset (0.03; 95% CI -0.01 to 0.07) or the subgroup (0.01; 95% CI -0.04 to 0.06). CONCLUSION: Utilization of a 2-zone, anterior-superior thoracic ultrasound protocol provided similar severity information to an 8-zone approach for a dataset of subjects with known or suspected heart failure.

Emergency Department Point-of-Care Echocardiography and Lung Ultrasound in Predicting COVID-19 Severity.

OBJECTIVES: We sought to determine if point-of-care ultrasound (POCUS) performed on patients with COVID-19 in the emergency department (ED) can help predict disease course, severity, or identify complications. METHODS: This was a retrospective cohort study of adult ED patients who tested positive for COVID-19 at hospital admission or within 2 weeks of presentation and received heart or lung POCUS. Clips were reviewed for presence of decreased left ventricular ejection fraction (LVEF), right ventricular dilation, presence of B-lines, and pleural line abnormalities. Patients with worsening hypoxemic respiratory failure or shock requiring higher level of care and patients who expired were considered to have developed severe COVID-19. Regression analysis was performed to determine if there was a correlation between ED POCUS findings and development of severe COVID-19. RESULTS: A total of 155 patients met study criteria; 148 patients had documented cardiac views and 116 patients had documented lung views (113 with both). Mean age was 66.5 years old (±18.6) and 53% of subjects were female. Subjects with decreased LVEF that was not previously documented had increased odds of having severe COVID during their hospitalization compared to those with old or no dysfunction (OR 5.66, 95% CI: 1.55-19.95, P = .08). The presence of pleural line abnormalities was also predictive for development of severe COVID (OR 2.68, 95% CI: 1.04-6.92, P = .04). CONCLUSION: POCUS findings of previously unidentified decreased LVEF and pleural line abnormalities in patients with COVID-19 evaluated in the ED were correlated to a more severe clinical course and worse prognosis.

Effect of Point-of-Care Ultrasound on Clinical Outcomes in Low-Resource Settings: A Systematic Review.

Despite the potential for improved patient care, little is known of the true effect of point-of-care ultrasound (POCUS) on patient outcomes in resource-limited settings. Electronic databases were searched using medical subject heading and free text terms related to POCUS and resource-limited settings through August 2020. Two authors independently selected studies, assessed methodological quality using the Downs and Black scale and extracted data. Twenty observational studies were included in the final review. All studies had moderate to high risk of bias. No studies exhibited an effect on the pre-specified primary outcome of mortality. Varying degrees of change in differential diagnosis and management were reported, but definitions varied widely among studies. Estimates for change in diagnosis as a result of POCUS ranged from 15% to 52%, and those for change in management, from 17% to 87%. Articles on POCUS clinical utility represent a small part (4.6%) of the scholastic literature dedicated to POCUS in low-resource settings. POCUS is a valuable intervention to consider in resource-limited settings, with the potential to change diagnosis and patient management. The exact magnitude of effect remains unknown. There is a continued need for large-scale experimental studies to investigate the effect of POCUS on patient diagnosis, management and mortality.

Interobserver Agreement and Correlation of an Automated Algorithm for B-Line Identification and Quantification With Expert Sonologist Review in a Handheld Ultrasound Device.

OBJECTIVES: B-lines are ultrasound artifacts that can be used to detect a variety of pathologic lung conditions. Computer-aided methods to detect and quantify B-lines may standardize quantification and improve diagnosis by novice users. We sought to test the performance of an automated algorithm for the detection and quantification of B-lines in a handheld ultrasound device (HHUD). METHODS: Ultrasound images were prospectively collected on adult emergency department patients with dyspnea. Images from the first 124 patients were used for algorithm development. Clips from 80 unique subjects for testing were randomly selected in a predefined proportion of B-lines (0 B-lines, 1-2 B-lines, 3 or more B-lines) and blindly reviewed by five experts using both a manual and reviewer-adjusted process. Intraclass correlation coefficient (ICC) and weighted kappa were used to measure agreement, while an a priori threshold of an ICC (3,k) of 0.75 and precision of 0.3 were used to define adequate performance. RESULTS: ICC between the algorithm and manual count was 0.84 (95% confidence interval [CI] 0.75-0.90), with a precision of 0.15. ICC between the reviewer-adjusted count and the algorithm count was 0.94 (95% CI 0.90-0.96), and the ICC between the manual and reviewer-adjusted counts was 0.94 (95% CI 0.90-0.96). Weighted kappa was 0.72 (95% CI 0.49-0.95), 0.88 (95% CI 0.74-1), and 0.85 (95% CI 0.89-0.96), respectively. CONCLUSIONS: This study demonstrates a high correlation between point-of-care ultrasound experts and an automated algorithm to identify and quantify B-lines using an HHUD. Future research may incorporate this HHUD in clinical studies in multiple settings and users of varying experience levels.

Paramedic-performed Prehospital Point-of-care Ultrasound for Patients with Undifferentiated Dyspnea: A Pilot Study.

INTRODUCTION: Thoracic ultrasound is frequently used in the emergency department (ED) to determine the etiology of dyspnea, yet its use is not widespread in the prehospital setting. We sought to investigate the feasibility and diagnostic performance of paramedic acquisition and assessment of thoracic ultrasound images in the prehospital environment, specifically for the detection of B-lines in congestive heart failure (CHF). METHODS: This was a prospective observational study of a convenience sample of adult patients with a chief complaint of dyspnea. Paramedics participated in a didactic and hands-on session instructing them how to use a portable ultrasound device. Paramedics assessed patients for the presence of B-lines. Sensitivity and specificity for the presence of bilateral B-lines and any B-lines were calculated based on discharge diagnosis. Clips archived to the ultrasound units were reviewed and paramedic interpretations were compared to expert sonologist interpretations. RESULTS: A total of 63 paramedics completed both didactic and hands-on training, and 22 performed ultrasounds in the field. There were 65 patients with B-line findings recorded and a discharge diagnosis for analysis. The presence of bilateral B-lines for diagnosis of CHF yielded a sensitivity of 80.0% (95% confidence interval [CI], 51.4-94.7%) and specificity of 72.0% (95% CI, 57.3-83.3), while presence of any B-lines was 93.3% sensitive (95% CI, 66.0-99.7%), and 50% specific (95% CI, 35.7-64.2%) for CHF. Paramedics archived 117 ultrasound clips of which 63% were determined to be adequate for interpretation. Comparison of paramedic and expert sonologist interpretation of images showed good inter-rater agreement for detection of any B-lines (k = 0.60; 95% CI, 0.36-0.84). CONCLUSION: This observational pilot study suggests that prehospital lung ultrasound for B-lines may aid in identifying or excluding CHF as a cause of dyspnea. The presence of bilateral B-lines as determined by paramedics is reasonably sensitive and specific for the diagnosis of CHF and pulmonary edema, while the absence of B lines is likely to exclude significant decompensated heart failure. The study was limited by being a convenience sample and highlighted some of the difficulties related to prehospital research. Larger funded trials will be needed to provide more definitive data.

The cost of waiting: Association of ED boarding with hospitalization costs.

BACKGROUND: Emergency Department (ED) boarding, the practice of holding patients in the ED after they have been admitted to the hospital due to unavailability of inpatient beds, is common and contributes to the public health crisis of ED crowding. Prior work has documented the harms of ED boarding on access and quality of care. Limited studies examine the relationship between ED boarding and an equally important domain of quality-the cost of care. This study evaluates the relationship between ED boarding, ED characteristics and risk-adjusted hospitalization costs utilizing national publicly-reported measures. METHODS: We conducted a cross-sectional analysis of two 2018 Centers for Medicare and Medicaid Services (CMS) Hospital Compare datasets: 1) Medicare Hospital Spending per Patient and 2) Timely and Effective Care. We constructed a hospital-level multivariate linear regression analysis to examine the association between ED boarding and Medicare spending per beneficiary (MSPB), adjusting for ED length of stay, door to diagnostic evaluation time, and ED patient volume. RESULTS: A total of 2903 hospitals were included in the analysis. ED boarding was significantly correlated with MSPB (r = 0.1774; p-value: < 0.0001). In multivariate regression, ED boarding was also positively associated with MSPB (Beta: 0.00015; p < 0.0001) after adjustment for other hospital level crowding indicators. CONCLUSION: We found a strong relationship between measures of ED crowding, including ED boarding, and risk-adjusted hospital spending. Future work should elucidate the mediators of this relationship. Policymakers and administrators should consider the financial harms of ED boarding when devising strategies to improve hospital care access and flow.

Automated Lung Ultrasound B-Line Assessment Using a Deep Learning Algorithm.

Shortness of breath is a major reason that patients present to the emergency department (ED) and point-of-care ultrasound (POCUS) has been shown to aid in diagnosis, particularly through evaluation for artifacts known as B-lines. B-line identification and quantification can be a challenging skill for novice ultrasound users, and experienced users could benefit from a more objective measure of quantification. We sought to develop and test a deep learning (DL) algorithm to quantify the assessment of B-lines in lung ultrasound. We utilized ultrasound clips ( n = 400 ) from an existing database of ED patients to provide training and test sets to develop and test the DL algorithm based on deep convolutional neural networks. Interpretations of the images by algorithm were compared to expert human interpretations on binary and severity (a scale of 0-4) classifications. Our model yielded a sensitivity of 93% (95% confidence interval (CI) 81%-98%) and a specificity of 96% (95% CI 84%-99%) for the presence or absence of B-lines compared to expert read, with a kappa of 0.88 (95% CI 0.79-0.97). Model to expert agreement for severity classification yielded a weighted kappa of 0.65 (95% CI 0.56-074). Overall, the DL algorithm performed well and could be integrated into an ultrasound system in order to help diagnose and track B-line severity. The algorithm is better at distinguishing the presence from the absence of B-lines but can also be successfully used to distinguish between B-line severity. Such methods could decrease variability and provide a standardized method for improved diagnosis and outcome.