Artificial intelligence-based classification of echocardiographic views.

Aims: Augmenting echocardiography with artificial intelligence would allow for automated assessment of routine parameters and identification of disease patterns not easily recognized otherwise. View classification is an essential first step before deep learning can be applied to the echocardiogram. Methods and results: We trained two- and three-dimensional convolutional neural networks (CNNs) using transthoracic echocardiographic (TTE) studies obtained from 909 patients to classify nine view categories (10 269 videos). Transthoracic echocardiographic studies from 229 patients were used in internal validation (2582 videos). Convolutional neural networks were tested on 100 patients with comprehensive TTE studies (where the two examples chosen by CNNs as most likely to represent a view were evaluated) and 408 patients with five view categories obtained via point-of-care ultrasound (POCUS). The overall accuracy of the two-dimensional CNN was 96.8%, and the averaged area under the curve (AUC) was 0.997 on the comprehensive TTE testing set; these numbers were 98.4% and 0.998, respectively, on the POCUS set. For the three-dimensional CNN, the accuracy and AUC were 96.3% and 0.998 for full TTE studies and 95.0% and 0.996 on POCUS videos, respectively. The positive predictive value, which defined correctly identified predicted views, was higher with two-dimensional rather than three-dimensional networks, exceeding 93% in apical, short-axis aortic valve, and parasternal long-axis left ventricle views. Conclusion: An automated view classifier utilizing CNNs was able to classify cardiac views obtained using TTE and POCUS with high accuracy. The view classifier will facilitate the application of deep learning to echocardiography.

Moderate Aortic Valve Stenosis Is Associated With Increased Mortality Rate and Lifetime Loss: Systematic Review and Meta-Analysis of Reconstructed Time-to-Event Data of 409 680 Patients.

BACKGROUND: The mortality risk attributable to moderate aortic stenosis (AS) remains incompletely characterized and has historically been underestimated. We aim to evaluate the association between moderate AS and all-cause death, comparing it with no/mild AS (in a general referral population and in patients with heart failure with reduced ejection fraction). METHODS AND RESULTS: A systematic review and pooled meta-analysis of Kaplan-Meier-derived reconstructed time-to-event data of studies published by June 2023 was conducted to evaluate survival outcomes among patients with moderate AS in comparison with individuals with no/mild AS. Ten studies were included, encompassing a total of 409 680 patients (11 527 with moderate AS and 398 153 with no/mild AS). In the overall population, the 15-year overall survival rate was 23.3% (95% CI, 19.1%-28.3%) in patients with moderate AS and 58.9% (95% CI, 58.1%-59.7%) in patients with no/mild aortic stenosis (hazard ratio [HR], 2.55 [95% CI, 2.46-2.64]; P<0.001). In patients with heart failure with reduced ejection fraction, the 10-year overall survival rate was 15.5% (95% CI, 10.0%-24.0%) in patients with moderate AS and 37.3% (95% CI, 36.2%-38.5%) in patients with no/mild AS (HR, 1.83 [95% CI, 1.69-2.0]; P<0.001). In both populations (overall and heart failure with reduced ejection fraction), these differences correspond to significant lifetime loss associated with moderate AS during follow-up (4.4 years, P<0.001; and 1.9 years, P<0.001, respectively). A consistent pattern of elevated mortality rate associated with moderate AS in sensitivity analyses of matched studies was observed. CONCLUSIONS: Moderate AS was associated with higher risk of death and lifetime loss compared with patients with no/mild AS.

Transgender Women Exhibit a Distinct Stress Echocardiography Profile Compared to Age-Matched Cisgender Counterparts: The Mayo Clinic Women's Heart Clinic Experience.

BACKGROUND: Stress echocardiography (SE) is an important modality in cardiovascular risk stratification and obstructive coronary artery disease (CAD) assessment. Binary sex-based parameters are classically used for interpretation of these studies, even among transgender women (TGW). CAD is a leading cause of morbidity and mortality in this population. Yet, it remains unclear whether TGW exhibit a distinct stress testing profile from their cisgender counterparts. METHODS: Using a matched case-control study design, we compared the echocardiographic stress testing profiles of TGW (n=43) with those of matched cisgender men (CGM, n=84) and cisgender women (CGW, n=86) at a single center. Relevant data, including demographics, comorbidities, and cardiac testing data were manually extracted from the patients' charts. RESULTS: The prevalence of hypertension and dyslipidemia was similar between TGW and CGW and lower than that of CGM (p= .003 and .009, respectively). The majority of comorbidities and lab values were similar. On average, TGW had higher heart rates than CGM (p=.002) and had lower blood pressures than CGM and CGW (p<.05). The TGW's double product and metabolic equivalents were similar to those among CGW and lower than those of CGM (p=.016, p=.018, respectively). On echocardiography, the left ventricular end-diastolic and end-systolic diameters among TGW were similar to those of CGW but lower than CGM's (p=.023, p=.018, respectively). Measures of systolic and diastolic function, except for the exercise mitral valve E:e' ratio which was lower in TGW than CGW (p=.029), were largely similar among the three groups. There was no difference in the wall motion score index, and therefore, no difference in the percentage of positive SE tests. CONCLUSION: Our study shows, for the first time, that TGW have a SE profile that is distinct from that of their cisgender counterparts. Larger, multicenter, prospective studies are warranted to further characterize the SE profile of TGW.

The Natural History of Atrial Functional Mitral Regurgitation.

BACKGROUND: The natural history of moderate/severe atrial functional mitral regurgitation (AFMR) is unknown. OBJECTIVES: The authors sought to study the incidence of left ventricular (LV) systolic dysfunction (LVSD), progression or regression of ≥mild-moderate AFMR, and impact on mortality. METHODS: Adults with left atrial (LA) volume index ≥40 mL/m2, ≥mild-moderate AFMR, and follow-up echocardiogram were followed for incident LVSD (ejection fraction <50% and ≥10% lower than baseline), progression of mild-moderate/moderate AFMR to severe, and persistent regression of AFMR to no/trivial. Relation of AFMR progression or regression as time-dependent covariates with all-cause mortality was studied. Incidence of LVSD was compared with patients with no/mild AFMR matched on age, sex, comorbidities and ejection fraction. Patients were followed until mitral intervention, myocardial infarction, or last follow-up. RESULTS: A total of 635 patients (median age 75 years, 51% female, 96% mild-moderate/moderate AFMR, 4% severe AFMR) were included. Over a median 2.2 years (Q1-Q3: 1.0-4.3 years), incidence rates per 100 person-years were 3.2 for LVSD (P = 0.52 vs patients with no/mild AFMR), 1.9 for progression of AFMR, and 3.9 for regression. Female sex and larger LA volume index were independently associated with progression, whereas younger age, male sex, absent atrial fibrillation, and higher LA emptying fraction were independently associated with regression. Neither AFMR progression nor regression was independently associated with mortality. Instead, independent risk factors for mortality included older age, concentric LV geometry, and higher estimated LV filling and pulmonary pressures. CONCLUSIONS: In patients with predominantly mild-moderate/moderate AFMR, regression of MR was more common than progression, but neither was associated with mortality. Instead, diastolic function abnormalities were more important. Over a median 2-year follow-up, LVSD risk was not increased.

Progression of Mild Mitral Annulus Calcification to Mitral Valve Dysfunction and Impact on Mortality.

BACKGROUND: Mitral annulus calcification (MAC) represents a degenerative process resulting in calcium deposition in the mitral valve apparatus. Mitral annulus calcification is associated with adverse clinical outcomes. We sought to examine the long-term significance of mild MAC and its relationship to subsequent mitral valve dysfunction (MVD) and mortality in patients without MVD on the initial echocardiogram. METHODS: A total of 1,420 patients with mild MAC and no MVD at baseline and 1 or more follow-up echocardiograms at least 1 year after the baseline echocardiogram were included in the analysis. For patients with >1 echocardiogram during follow-up, the last echocardiogram was used. The same criteria were used to identify 6,496 patients without MAC. Mitral valve dysfunction was defined as mitral regurgitation (MR) and/or mitral stenosis (MS) of moderate or greater severity. Mixed disease was defined as the concurrent presence of both moderate or greater MS and MR. The primary end point was development of MVD, and the secondary end point was all-cause mortality. RESULTS: For patients with mild MAC, age was 74 ± 10 years and 528 (37%) were female. Over a median follow-up of 4.7 (interquartile range, 2.7-6.9) years, 215 patients with mild MAC developed MVD, including MR in 170 (79%), MS in 37 (17%), and mixed disease in 8 (4%). In a multivariable regression model compared to patients without MAC, the presence of mild MAC was independently associated with increased mortality (hazard ratio = 1.43; 95% CI 1.24, 1.66; P < .001). Kaplan-Meier 4-year survival rates were 80% and 90% for patients with mild MAC and no MAC, respectively. CONCLUSIONS: Mild MAC observed on transthoracic echocardiography is an important clinical finding with prognostic implications for both valvular function and mortality.

Coronary Artery Tortuosity and Spontaneous Coronary Artery Dissection: Association With Echocardiography and Global Longitudinal Strain, Fibromuscular Dysplasia, and Outcomes.

BACKGROUND: The etiology and significance of coronary artery tortuosity (TCA) among patients with spontaneous coronary artery dissection (SCAD) are unknown. The aim of this prospective imaging cohort study was to report echocardiographic findings and evaluate whether TCA correlates with cardiac anatomy and function among patients with SCAD. Comorbidities including fibromuscular dysplasia (FMD) and outcomes were also assessed. METHODS: TCA was determined on coronary angiography performed during the diagnosis of SCAD, and cardiac structure and function were evaluated using prospective comprehensive echocardiography. RESULTS: Among 116 patients with SCAD, the mean age at echocardiography was 50.8 ± 8.8 years, a median of 10.9 months after SCAD. Sixty-two patients (53.4%) had FMD, 41 (35.3%) had histories of hypertension, and 17 (14.8%) were hypertensive during echocardiography. Most patients (n = 78 [69%]) had normal left ventricular geometry with normal median ejection fraction (61%; interquartile range, 56% to 64%) and normal global longitudinal strain (-22.2%; interquartile range, -24.0% to -19.9%). Fifteen patients (13.4%) had diastolic dysfunction that was associated with hypertension at the time of echocardiography. Patients with TCA (n = 96 [82.8%]) were older (mean age, 52.1 ± 8.0 vs 44.7 ± 9.9 years; P < .001) with a higher prevalence of FMD (59.4% vs 25%, P = .007) but a similar prevalence of hypertension (35% vs 35%, P > .99) compared with patients without TCA. Across the age range (31.5 to 66.9 years), each decade of age was associated with an approximately 0.89-unit increase in coronary tortuosity score (P < .0001). Echocardiographic parameters were not significantly different between the two groups. Median follow-up duration was 4.4 years (95% CI, 3.8 to 5.2 years). The Kaplan-Meier 3-year SCAD recurrence rate was 9.4% (95% CI, 3.7% to 14.8%). There were no deaths. CONCLUSIONS: The majority of patients with SCAD had normal or near normal echocardiographic results, including global longitudinal strain, with no differences according to TCA. However, patients with SCAD with TCA were older, with a higher prevalence of FMD.

A Pragmatic Study of Cardiovascular Disease During Long-Term COVID-19.

BACKGROUND: Many patients diagnosed with COVID-19 have persistent cardiovascular symptoms, but whether this represents a true cardiac process is unclear. This study assessed whether symptoms associated with long COVID among patients referred for cardiovascular evaluation are associated with objective abnormalities on cardiac testing to explain their clinical presentation. METHODS: A retrospective cohort study of 40,462 unique patients diagnosed with COVID-19 at our tertiary referral was conducted and identified 363 patients with persistent cardiovascular symptoms a minimum of 4 weeks after polymerase chain reaction confirmed COVID-19 infection. Patients had no cardiovascular symptoms prior to COVID-19 infection. Each patient was referred for cardiovascular evaluation at a tertiary referral center. The incidence and etiology of abnormalities on cardiovascular testing among patients with long COVID symptoms are reported here. The cohort was subsequently divided into 3 categories based on the dominant circulating severe acute respiratory syndrome coronavirus 2 variant at the time of initial infection for further analysis. RESULTS: Among 40,462 unique patients diagnosed with COVID-19 at our tertiary referral center from April 2020 to March 2022, 363 (0.9%) patients with long COVID were evaluated by Cardiology for possible cardiac sequelae from COVID and formed the main study cohort. Of these, 229 (63%) were vaccinated and 47 (12.9%) had severe initial infection, receiving inpatient treatment for COVID prior to developing long COVID symptoms. Symptoms were associated with a cardiac cause in 85 (23.4%), of which 52 (14.3%) were attributed to COVID; 39 (10.7%) with new cardiac disease from COVID, and 13 (3.6%) to worsening of pre-existing cardiac disease after COVID infection. The median troponin change in 45 patients with troponin measurements within 4 weeks of acute infection was +4 ng/dL (9 to 13 ng/dL). Among the total cohort with long COVID, 83.7% were diagnosed during the pre-Delta phase, 13.2% during the Delta phase, and 3.1% during the Omicron phase of the pandemic. There were 6 cases of myocarditis, 11 rhythm disorders, 8 cases of pericarditis, 5 suspected cases of endothelial dysfunction, and 33 cases of autonomic dysfunction. CONCLUSION: This pragmatic retrospective cohort study suggests that patients with long COVID referred for cardiovascular evaluation infrequently have new, objective cardiovascular disease to explain their clinical presentation. A multidisciplinary, patient-centered approach is warranted for symptom management along with conservative use of diagnostic testing.

High Resting Coronary Flow Velocity by Echocardiography Is Associated With Worse Survival in Patients With Chronic Coronary Syndromes.

BACKGROUND: Resting coronary flow velocity (CFV) in the mid-distal left anterior descending coronary artery can be easily assessed with transthoracic echocardiography. In this observational study, the authors sought to assess the relationship between resting CFV, CFV reserve (CFVR), and outcome in patients with chronic coronary syndromes. METHODS AND RESULTS: In a prospective multicenter study design, the authors retrospectively analyzed 7576 patients (age, 66±11 years; 4312 men) with chronic coronary syndromes and left ventricular ejection fraction ≥50% referred for dipyridamole stress echocardiography. Recruitment (years 2003-2021) involved 7 accredited laboratories, with interobserver variability <10% for CFV measurement at study entry. Baseline peak diastolic CFV was obtained by pulsed-wave Doppler in the mid-distal left anterior descending coronary artery. CFVR (abnormal value ≤2.0) was assessed with dipyridamole. All-cause death was the only end point. The mean CFV of the left anterior descending coronary artery was 31±12 cm/s. The mean CFVR was 2.32±0.60. During a median follow-up of 5.9±4.3 years, 1121 (15%) patients died. At multivariable analysis, resting CFV ≥32 cm/s was identified by a receiver operating curve as the best cutoff and was independently associated with mortality (hazard ratio [HR], 1.24 [95% CI, 1.10-1.40]; P<0.0001) together with CFVR ≤2.0 (HR, 1.78 [95% CI, 1.57-2.02]; P<0.0001), age, diabetes, history of coronary surgery, and left ventricular ejection fraction. When both CFV and CFVR were considered, the mortality rate was highest in patients with resting CFV ≥32 cm/s and CFVR ≤2.0 and lowest in patients with resting CFV <32 cm/s and CFVR >2.0. CONCLUSIONS: High resting CFV is associated with worse survival in patients with chronic coronary syndromes and left ventricular ejection fraction ≥50%. The value is independent and additive to CFVR. The combination of high resting CFV and low CFVR is associated with the worst survival.

Machine Learning for Diagnosis of Pulmonary Hypertension by Echocardiography.

OBJECTIVE: To evaluate a machine learning (ML)-based model for pulmonary hypertension (PH) prediction using measurements and impressions made during echocardiography. METHODS: A total of 7853 consecutive patients with right-sided heart catheterization and transthoracic echocardiography performed within 1 week from January 1, 2012, through December 31, 2019, were included. The data were split into training (n=5024 [64%]), validation (n=1275 [16%]), and testing (n=1554 [20%]). A gradient boosting machine with enumerated grid search for optimization was selected to allow missing data in the boosted trees without imputation. The training target was PH, defined by right-sided heart catheterization as mean pulmonary artery pressure above 20 mm Hg; model performance was maximized relative to area under the receiver operating characteristic curve using 5-fold cross-validation. RESULTS: Cohort age was 64±14 years; 3467 (44%) were female, and 81% (6323/7853) had PH. The final trained model included 19 characteristics, measurements, or impressions derived from the echocardiogram. In the testing data, the model had high discrimination for the detection of PH (area under the receiver operating characteristic curve, 0.83; 95% CI, 0.80 to 0.85). The model's accuracy, sensitivity, positive predictive value, and negative predictive value were 82% (1267/1554), 88% (1098/1242), 89% (1098/1241), and 54% (169/313), respectively. CONCLUSION: By use of ML, PH could be predicted on the basis of clinical and echocardiographic variables, without tricuspid regurgitation velocity. Machine learning methods appear promising for identifying patients with low likelihood of PH.

Mitral Annulus Calcium Score in Patients With Calcific Mitral Stenosis Undergoing Invasive Hemodynamic Assessment.

BACKGROUND: Aortic valve calcium score is associated with hemodynamic severity of aortic stenosis. Whether this association is present in calcific mitral stenosis remains unknown. METHODS AND RESULTS: This study was a retrospective analysis of consecutive patients with mitral stenosis secondary to mitral annular calcification (MAC) undergoing transseptal catheterization. All patients underwent invasive mitral valve assessment via direct left atrial and left ventricular pressure measurement. Computed tomography within 1 year of cardiac catheterization and with adequate visualization of the mitral annulus was included. MAC calcium score quantification by Agatston method was obtained offline using dedicated software (Aquarius, TeraRecon, V.4). Median patient age was 66.9±11.2 years, 47% of patients were women, 50% had coronary artery disease, 40% had atrial fibrillation, 47% had prior cardiac surgery, and 33% had prior chest radiation. Median diastolic mitral valve gradient was 9.4±3.4 mm Hg on echocardiography and 8.5±4 mm Hg invasively. Invasive median mitral valve area using the Gorlin formula was 1.87±0.9 cm2. Median MAC calcium score for the cohort was 7280±7937 Hounsfield units. MAC calcium score correlated with the presence of atrial fibrillation (P=0.02) but was not associated with other comorbidities. There was no correlation between MAC calcium score and mitral valve area (r=0.07; P=0.6) or mitral valve gradient (r=-0.03; P=0.8). CONCLUSIONS: MAC calcium score did not correlate with invasively measured mitral valve gradient and mitral valve area in patients with MAC-related mitral stenosis, suggesting that calcium score should not be used as a surrogate for invasive hemodynamic parameters.

Biatrial myopathy in heart failure with preserved ejection fraction.

AIM: Left atrial (LA) myopathy is increasingly recognized as an important phenotypic trait in heart failure (HF) with preserved ejection fraction (HFpEF). Right atrial (RA) remodelling and dysfunction also develop in HFpEF, but little data are available regarding the clinical characteristics and pathophysiology among patients with isolated LA, RA, or biatrial myopathy. METHODS AND RESULTS: Patients with HFpEF underwent invasive haemodynamic exercise testing, comprehensive imaging including speckle tracking strain echocardiography, and clinical follow-up at Mayo Clinic between 2006 and 2018. LA myopathy was defined as LA volume index >34 ml/m2 and/or LA reservoir strain ≤24% and RA myopathy by RA volume index >39 ml/m2 in men and >33 ml/m2 in women and/or RA reservoir strain ≤19.8%. Of 476 consecutively evaluated patients with HFpEF defined by invasive exercise testing with evaluable atrial structure/function, 125 (26%) had no atrial myopathy, 147 (31%) had isolated LA myopathy, 184 (39%) had biatrial myopathy, and 20 (4%) had isolated RA myopathy. Patients with HFpEF and biatrial myopathy had more atrial fibrillation, poorer left ventricular systolic and diastolic function, more severe pulmonary vascular disease, tricuspid regurgitation, ventricular interdependence and right ventricular dysfunction, and poorer cardiac output reserve with exercise. There were 94 patients with events over a median follow-up of 2.9 (interquartile range 1.4-4.6) years. Individuals with biatrial myopathy had an 84% higher risk of HF hospitalization or death as compared to those with isolated LA myopathy (hazard ratio 1.84; 95% confidence interval 1.16-2.92, p = 0.01). CONCLUSIONS: Biatrial myopathy identifies patients with more advanced HFpEF characterized by more severe pulmonary vascular disease, right HF, poorer cardiac reserve, and a greater risk for adverse outcomes. Further study is required to define optimal strategies to treat and prevent biatrial myopathy in HFpEF.

Global Longitudinal Strain as Predictor of Inducible Ischemia in No Obstructive Coronary Artery Disease in the CIAO-ISCHEMIA Study.

BACKGROUND: Global longitudinal strain (GLS) is a sensitive marker for identifying subclinical myocardial dysfunction in obstructive coronary artery disease (CAD). Little is known about the relationship between GLS and ischemia in patients with myocardial ischemia and no obstructive CAD (INOCA). OBJECTIVES: To investigate the relationship between resting GLS and ischemia on stress echocardiography (SE) in patients with INOCA. METHODS: Left ventricular GLS was calculated offline on resting SE images at enrollment (n = 144) and 1-year follow-up (n = 120) in the CIAO-ISCHEMIA (Changes in Ischemia and Angina over One year in International Study of Comparative Health Effectiveness with Medical and Invasive Approaches trial screen failures with no obstructive CAD on computed tomography [CT] angiography) study, which enrolled participants with moderate or severe ischemia by local SE interpretation (≥3 segments with new or worsening wall motion abnormality and no obstructive (<50% stenosis) on coronary computed tomography angiography. RESULTS: Global longitudinal strain values were normal in 83.3% at enrollment and 94.2% at follow-up. Global longitudinal strain values were not associated with a positive SE at enrollment (GLS = -21.5% positive SE vs GLS = -19.9% negative SE, P = .443) or follow-up (GLS = -23.2% positive SE vs GLS = -23.1% negative SE, P = .859). Significant change in GLS was not associated with positive SE in follow-up (P = .401). Regional strain was not associated with colocalizing ischemia at enrollment or follow-up. Changes in GLS and number of ischemic segments from enrollment to follow-up showed a modest but not clinically meaningful correlation (ß = 0.41; 95% CI, 0.16, 0.67; P = .002). CONCLUSIONS: In this cohort of INOCA patients, resting GLS values were largely normal and did not associate with the presence, severity, or location of stress-induced ischemia. These findings may suggest the absence of subclinical myocardial dysfunction detectable by echocardiographic strain analysis at rest in INOCA.

Prognostic Significance of Elevated Left Ventricular Filling Pressures with Exercise: Insights from a Cohort of 14,338 Patients.

BACKGROUND: Exercise echocardiography can assess for cardiovascular causes of dyspnea other than coronary artery disease. However, the prevalence and prognostic significance of elevated left ventricular (LV) filling pressures with exercise is understudied. METHODS: We evaluated 14,338 patients referred for maximal symptom-limited treadmill echocardiography. In addition to assessment of LV regional wall motion abnormalities (RWMAs), we measured patients' early diastolic mitral inflow (E), septal mitral annulus relaxation (e'), and peak tricuspid regurgitation velocity before and immediately after exercise. RESULTS: Over a mean follow-up of 3.3 ± 3.4 years, patients with E/e' ≥15 with exercise (n = 1,323; 9.2%) had lower exercise capacity (7.3 ± 2.1 vs 9.1 ± 2.4 metabolic equivalents, P < .0001) and were more likely to have resting or inducible RWMAs (38% vs 18%, P < .0001). Approximately 6% (n = 837) had elevated LV filling pressures without RWMAs. Patients with a poststress E/e' ≥15 had a 2.71-fold increased mortality rate (2.28-3.21, P < .0001) compared with those with poststress E/e' ≤ 8. Those with an E/e' of 9 to 14, while at lower risk than the E/e' ≥15 cohort (hazard ratio [HR] = 0.58 [0.48-0.69]; P < .0001), had higher risk than if E/e' ≤8 (HR = 1.56 [1.37-1.78], P < .0001). On multivariable analysis, adjusting for age, sex, exercise capacity, LV ejection fraction, and presence of pulmonary hypertension with stress, patients with E/e' ≥15 had a 1.39-fold (95% CI, 1.18-1.65, P < .0001) increased risk of all-cause mortality compared with patients without elevated LV filling pressures. Compared with patients with E/e' ≤ 15 after exercise, patients with E/e' ≤15 at rest but elevated after exercise had a higher risk of cardiovascular death (HR = 8.99 [4.7-17.3], P < .0001). CONCLUSION: Patients with elevated LV filling pressures are at increased risk of death, irrespective of myocardial ischemia or LV systolic dysfunction. These findings support the routine incorporation of LV filling pressure assessment, both before and immediately following stress, into the evaluation of patients referred for exercise echocardiography.