Citations

Abstract 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.

AbstractChronic lung disease (CLD) is the second leading cause of pulmonary hypertension (PH) and is associated with significant morbidity and mortality. Although PH associated with CLD (PH‐CLD) leads to impaired health‐related quality of life (HRQOL), there are no validated tools to assess HRQOL in PH‐CLD. The Pulmonary Arterial Hypertension–Symptoms and Impact Questionnaire (PAH‐SYMPACT) is an HRQOL instrument aimed at assessing the symptoms and impact of PH on overall function and well‐being. We performed a single‐center prospective cohort study using PAH‐SYMPACT scores to compare symptoms, exercise capacity and HRQOL in patients with PAH and PH‐CLD. One hundred and twenty‐five patients (99 patients with idiopathic/heritable PAH and 26 with PH‐CLD) completed the PAH‐SYMPACT questionnaire which consists of 22 questions that assess HRQOL across four domains: cardiopulmonary (CP) symptoms, cardiovascular (CV) symptoms, physical impact (PI), and cognitive/emotional (CE) impact. Higher scores indicate worse HRQOL. We compared patients with PAH and PH‐CLD using a Wilcoxon rank sum or chi‐squared test as appropriate. Multivariate linear regression analysis was used to assess the relationship between PH classification and SYMPACT scores. Compared to PAH, patients with PH‐CLD were older, more likely to use oxygen and had worse functional class and exercise capacity. While there was no significant difference between the two groups in CP, CV, or CE domain scores, patients with PH‐CLD had significantly worse PI scores by univariate (1.79 vs. 1.13, p < 0.001) and multivariate analysis (1.61 vs. 1.17, p = 0.02) and overall worse SYMPACT scores (1.19 vs. 0.91, p = 0.03). In conclusion, patients with PH‐CLD have worse HRQOL as assessed by the PAH‐SYMPACT questionnaire versus patients with PAH. Although PAH‐SYMPACT has not been validated in PH‐CLD, the results of this study can guide clinicians in understanding the symptoms and impact of PH‐CLD relative to PAH.

Background: Enlargement or dilation of the left ventricle (LV) is associated with increased cardiovascular morbidity and mortality, and is an established precursor of ventricular dysfunction and heart failure in almost all major cardiovascular conditions. Although there are existing artificial intelligence frameworks that make use of deep learning to detect LV enlargement, many entail LV segmentation, which can be error-prone especially for point-of-care ultrasound (POCUS) images. Goal: To develop a fully end-to-end deep learning framework that takes as input either transthoracic echocardiographic (TTE) or POCUS 2D image clips, and classifies directly (i.e., without first segmenting the LV) whether a patient has an enlarged LV. Methods: Using a sample of retrospective TTE data enriched for LV enlargement (out of 13,868 patients total, 3,351 had an enlarged LV defined using sex-specific guideline-directed criteria), we trained a classifier that takes as input apical 2-chamber and parasternal long axis videos, and provides as output a binary label denoting whether a patient has an enlarged LV. Then, we evaluated the trained model using a holdout testing dataset of retrospective TTE data (2,087 patients) as well as data from a prospective cohort of 428 patients from whom TTE and POCUS were collected simultaneously. Results: For the retrospective TTE testing dataset, the AUC was 0.851 (95% confidence interval (CI) of 0.830-871). For the prospective cohort, the model performed comparably for both TTE and POCUS video clips (TTE: AUC of 0.865, 95% CI of 0.823-0.906; POCUS: AUC of 0.840, 95% CI of 0.797-0.883). Conclusions: The current deep learning model performed comparably well for both TTE and POCUS echocardiograms. We regard these results as a proof-of-concept of the potential utility of this model in helping point-of-care clinicians make decisions regarding whether patients should receive further evaluation for LV enlargement and associated conditions.

Background: Diagnosis of heart failure with preserved ejection fraction (HFpEF) is challenging due to non-diagnostic imaging outputs and discordant clinical features. Using only a single videoclip from a standard transthoracic echocardiogram (TTE), we developed an artificial intelligence (AI) model to differentiate patients with HFpEF from those without HFpEF. In a separate test population, we assessed the model’s ability to predict cardiac mortality. Methods: A 3-dimensional convolutional neural network was developed and trained on apical 4-chamber videoclips to classify patients with HFpEF (cases) versus without HFpEF (controls). Overall and cardiac mortality were obtained from the National Death Index. Results: The model was independently tested in 646 cases and 638 controls from 8 centers in 4 states. Sensitivity (87.8%; 84.5-90.9%) and specificity (81.9%; 78.2-85.6%) were high and sustained in various subgroups, including increased left atrial volume index, average E/e’ ≥15, and pulmonary artery systolic pressure >35 mm Hg. During follow-up (median [IQR] 3.4 [1.7-6.5] years), 540 patients died. There were 135 cardiac deaths attributable to heart failure (47%), coronary artery disease (41%), valve disease (4%), arrhythmia (4%), and others (5%). Compared to a negative output, all-cause mortality was higher in patients with a positive (HR 2.9 [2.4-3.5]) or uncertain output (HR 1.8 [1.2-2.5]), and cardiac mortality (Figure) was also higher in patients with a positive (HR 7.7 [4.5-13.3]) or uncertain output (HR 3.4 [1.4-8.1]). Cardiac mortality differences according to AI model prediction remained after age adjustment (positive: HR 5.9 [3.3-10.4], p<0.001; uncertain: HR 2.8 [1.2-6.8], p=0.02). Conclusion: A novel AI model to detect HFpEF based on a single routinely acquired TTE videoclip also identifies those at increased risk of cardiac mortality.

Background: Recently, our team developed an end-to-end artificial intelligence (AI) framework that can successfully estimate left ventricular ejection fraction (LVEF) from echocardiogram videos. However, this framework requires specific views (A2C and PLAX), and cannot estimate LVEF for studies from which these views are missing. Goal: To develop a deep learning model to estimate LVEF from any type and number of echocardiogram video(s). Methods: We built a deep learning model for B-mode echocardiogram videos of the left ventricle (A2C, A3C, A4C, or PLAX views), using retrospective transthoracic echocardiography (TTE) data. The model was evaluated with two test datasets: one with random valid (i.e., showing the left ventricle) views, and one with preselected views (both having 433 patients). Later, the model was tested with a prospective patient cohort that had both TTE and point-of-care ultrasound (POCUS) data collected simultaneously (393 patients). Results: On the retrospective TTE test set, we observed excellent performance (five random valid views per study: RMSE of 6.87% and Pearson correlation (PC) of 0.82 on LVEF estimation, and an AUC of 0.96 (95% CI: of 0.94-0.98) on low LVEF classification (<=40%); preselected views: RMSE of 6.75%, PC of 0.83, and an AUC of 0.96 (95% CI: 0.94-0.98)). For both TTE and POCUS data from the prospective cohort, we also observed strong performance (TTE: RMSE of 5.62%, PC of 0.87, and an AUC of 0.97; POCUS: RMSE of 6.12%, PC of 0.84, and an AUC of 0.96). For the prospective test sets as well as for the retrospective TTE test set with preselected views, results reflect the estimated LVEF average of A2C, A3C, A4C, and PLAX videos from each patient. Conclusions: For both TTE and POCUS data, the current model showed strong performance which was comparable to that of our previously established view-specific model. Alongside this comparable performance, the current model surpasses our previous model in terms of flexibility, as it isn't dependent on view classification and allows a broader selection in terms of the number and types of videos per study or patient. This innovative approach could prove particularly beneficial in point-of-care situations in which it may not be feasible to obtain high-quality, view-specific videos.

Background: Clinically significant tricuspid regurgitation (TR) is an important contributor to morbidity and mortality. The Tricuspid Regurgitation Impact on Outcomes (TRIO) score based on 8 parameters (age, gender, chronic heart failure, chronic lung disease, heart rate, creatinine, aspartate transaminase level, and TR severity) is used for risk stratification. The performance of the TRIO score in patients with TR secondary to severe pulmonary hypertension (PH) is unknown. Research Questions: To evaluate whether the TRIO score can be applied in patients with TR and pulmonary hypertension (PH), and to identify other predictors of mortality. Methods: Retrospective study including patients with TR ≥ moderate. Severe PH was defined as right ventricular systolic pressure (RVSP) ≥50 mmHg. The impact of clinical, echocardiographic and laboratory variables on mortality was evaluated with the Cox proportional hazard method. Results: A total of 13,482 patients from the TRIO population were included; severe PH was present in 6,530 (48.4%); mean age 73.4 + 13.3, 55.7% females. TRIO score performance was modest (AUC 0.62) in risk stratification (Figure A). PH patients had significantly higher mortality than those without PH at all levels of TRIO score (Figure B); differences were progressively lower with increasing TRIO scores. At multivariate analysis, only TRIO score [HR = 1.54 (1.35-1.77) per unit change], presence of an elevated NT-proBNP [HR = 1.02 (1.02-1.03) per 1000 pg/ml change], RVSP (HR = 1.08 (1.02-1.16) per 10 mmHg change], presence of RV dysfunction [HR = 1.15, (1.05-1.26)] and pulmonary artery pulsatility index (PAPi) ≤ 4 [HR = 1.34 (1.07-1.67)] were significant predictors of mortality (model AUC 0.65). Conclusions: TRIO score allows risk stratification in PH patients, but performance is modest; incorporating additional factors marginally increased model performance. PH patients have higher mortality at all levels of TRIO score, more so at low scores.

Background: Atrial fibrillation (a-fib) occurs in up to 40% of patients after noncardiac surgery and is associated with increased risk of adverse outcomes. Practice guidelines recommend stress testing, including stress echocardiography (SE) to assess for coronary artery disease in selected high-risk patients prior to noncardiac surgery. The utility of using SE findings for predicting postoperative a-fib is unknown. Hypothesis: Myocardial ischemia can contribute to arrhythmogenic potential; other echocardiographic variables are likely associated with a-fib. Findings on SE may identify patients at risk for a-fib after noncardiac surgery. Aims: To determine if certain SE findings predict a-fib following noncardiac surgery. Methods: The study included patients undergoing pharmacological and exercise SE before noncardiac surgery. The primary outcome was a-fib/flutter within 30 days after surgery. Multivariate analysis was adjusted for age, cardiovascular comorbidities, and markers of surgical risk. Results: This analysis included 2,376 patients (mean age 68 +/- 7 years; 60% male); 1,986 patients had no known history of a-fib/flutter. Of these, 48 (2.4%) developed postoperative a-fib/flutter. In multivariate analysis, SE variables associated with development of new a-fib/flutter included left ventricular dilation (OR 2.11, p=0.049), increased left atrial volume (OR 1.19, p=0.045), and increased left ventricular mass index (OR 1.13, p=0.043) on rest images. The presence of ischemia or infarct on stress echocardiogram was not associated with postoperative a-fib/flutter. A total of 390 patients had a history of paroxysmal a-fib/flutter and were in sinus rhythm at the time of surgery. Of these, 90 (23%) experienced recurrent a-fib/flutter within 30 days of surgery. Multivariate analysis did not reveal any SE variables significantly associated with recurrence, including presence of myocardial ischemia or infarct at the time of SE. Conclusions: All SE variables associated with postoperative a-fib following noncardiac surgery are available from resting images alone. Therefore, SE that includes comprehensive resting images may offer benefit in the pre-operative evaluation of selected patients.

Objective Tricuspid regurgitation (TR) is a prevalent valve disease associated with significant morbidity and mortality. We aimed to apply machine learning (ML) to assess risk stratification in patients with ≥moderate TR. Methods Patients with ≥moderate TR on echocardiogram between January 2005 and December 2016 were retrospectively included. We used 70% of data to train ML-based survival models including 27 clinical and echocardiographic features to predict mortality over a 3-year period on an independent test set (30%). To account for differences in baseline comorbidities, prediction was performed in groups stratified by increasing Charlson Comorbidity Index (CCI). Permutation feature importance was calculated using the best-performing model separately in these groups. Results Of 13 312 patients, mean age 72 ± 13 years and 7406 (55%) women, 7409 (56%) had moderate, 2646 (20%) had moderate–severe and 3257 (24%) had severe TR. The overall performance for 1-year mortality by 3 ML models was good, c-statistic 0.74–0.75. Interestingly, performance varied between CCI groups, (c-statistic = 0.774 in lowest CCI group and 0.661 in highest CCI group). The performance decreased over 3-year follow-up (average c-index 0.78). Furthermore, the top 10 features contributing to these predictions varied slightly with the CCI group, the top features included heart rate, right ventricular systolic pressure, blood pressure, diuretic use and age. Conclusions Machine learning of common clinical and echocardiographic features can evaluate mortality risk in patients with TR. Further refinement of models and validation in prospective studies are needed before incorporation into the clinical practice.

Abstract Background Pulmonary veno-occlusive disease (PVOD) is a rare cause of pulmonary hypertension that primarily results in intimal fibrosis of the pulmonary venules and veins, the venous lesions are often associated with changes in interlobular septa, lung parenchyma, and pulmonary arterioles and arteries. Purpose We sought to study the clinical characteristics of patients with confirmed PVOD at a single large referral center. Methods We retrospectively reviewed data from a total of 675 patients with pulmonary hypertension (identified by ICD codes or search of text) from 2017 and 2022 from a database obtained from electronic medical records at our institution. Patients with group 2 pulmonary hypertension were excluded. Only 29 cases were confirmed by histology or genetic testing. Results Out of the 29 patients confirmed with PVOD, 28 (96.5%) had a characteristic histology (lung biopsy (55.1%), explanted lung for transplant (27.5%) and autopsy (13.7%)) and 1 (3.4%) had a positive genetic test for EIF2AK4 biallelic mutation. The age at diagnosis was 53±16 years, 14 (48.2%) were men, 2 (6.8%) had systemic scleroderma, 2 (6.8%) had psoriasis, 2 (6.8%) were exposed to cytotoxic agents, 16 (55.1%) were smokers and 2 (6.8%) had a family history of pulmonary hypertension. All patients had computed tomography of the chest but only 12 (41.3%) had all 3 typical features found in PVOD (ground glass opacities, interlobular septal thickening, and lymphadenopathy). The most common feature was interlobular septal thickening in 26 (89.6%), followed by ground glass opacities present in 25 (86.2%) and lymphadenopathy in 12 (41.3%) of the patients. At diagnosis, mean pulmonary artery pressure was 45±15 mmHg and pulmonary vascular resistance was 9±6 WU. The median DLCO was 41 (34-58) %. On echocardiogram, 14 (48.2%) had severely enlarged right ventricle (RV), RV function was ≥moderately reduced in 17 (58.6%), RV systolic pressure was 63±20 mmHg and 6 (20.6%) had pericardial effusion. Out of 21 (72.4%) patients who received pulmonary vasodilators, 4 (13.7%) received single agent. 10 (34.4%) received 2 agents, 6 (20.6%) 3 agents, and 1 (3.4%) received 4 agents. 35.7% of these treated patients reported an improvement or initial worsening of symptoms with subsequent improvement afterwards. 18 (62%) patients with PVOD died, 12 (41.3%) had a lung transplant and 6 (20.6%) were still alive without the transplant until the day data was collected. Mean duration from diagnosis to death and lung transplantation was 446 ± 405 days and 274±479 days respectively. Conclusion PVOD is a rare and lethal cause of pulmonary hypertension. When suspected, a trial of medical therapy should be initiated carefully. The evaluation for lung transplant should be considered early given its high mortality.