Phenotypic Clustering of Left Ventricular Diastolic Function Parameters: Patterns and Prognostic Relevance.

OBJECTIVES: This study sought to explore the natural clustering of echocardiographic variables used for assessing left ventricular (LV) diastolic dysfunction (DD) in order to isolate high-risk phenotypic patterns and assess their prognostic significance. BACKGROUND: Assessment of LV DD is important in the management and prognosis of cardiovascular diseases. Data-driven approaches such as cluster analysis may be useful in segregating similar cases without the constraint of an a priori algorithm for risk stratification. METHODS: The study included a convenience sample of 866 consecutive patients referred for myocardial function assessment (age 65 ± 17 years; 55.3% women; ejection fraction 60 ± 9%) for whom echocardiographic parameters of DD assessment were obtained per conventional guideline recommendations. Unsupervised, hierarchical cluster analysis of these parameters was conducted using the Ward linkage method. Major adverse cardiovascular events, hospitalization, and mortality were compared between conventional and cluster-based classifications. RESULTS: Clustering algorithms for screening the presence of DD in 559 of 866 patients identified 2 distinct groups and revealed modest agreement with conventional classification (kappa = 0.41, p < 0.001). Further cluster analysis in 387 patients with DD helped to classify the severity of DD into 2 groups, with good agreement with conventional classification (kappa = 0.619, p < 0.001). Survival analyses of patients assessed by both clustering algorithms for screening and grading DD showed improved prediction of event-free survival by clusters over conventional classification for all-cause mortality and cardiac mortality, even after accounting for a multivariable, balanced propensity score. CONCLUSIONS: An unsupervised assessment of echocardiographic variables for assessing LV DD revealed unique patterns of grouping. These natural patterns of clustering may better identify patient groups who have similar risk, and their incorporation into clinical practice may help eliminate indeterminate results and improve clinical outcome prediction.

Machine-Learning Algorithms to Automate Morphological and Functional Assessments in 2D Echocardiography.

BACKGROUND: Machine-learning models may aid cardiac phenotypic recognition by using features of cardiac tissue deformation. OBJECTIVES: This study investigated the diagnostic value of a machine-learning framework that incorporates speckle-tracking echocardiographic data for automated discrimination of hypertrophic cardiomyopathy (HCM) from physiological hypertrophy seen in athletes (ATH). METHODS: Expert-annotated speckle-tracking echocardiographic datasets obtained from 77 ATH and 62 HCM patients were used for developing an automated system. An ensemble machine-learning model with 3 different machine-learning algorithms (support vector machines, random forests, and artificial neural networks) was developed and a majority voting method was used for conclusive predictions with further K-fold cross-validation. RESULTS: Feature selection using an information gain (IG) algorithm revealed that volume was the best predictor for differentiating between HCM ands. ATH (IG = 0.24) followed by mid-left ventricular segmental (IG = 0.134) and average longitudinal strain (IG = 0.131). The ensemble machine-learning model showed increased sensitivity and specificity compared with early-to-late diastolic transmitral velocity ratio (p < 0.01), average early diastolic tissue velocity (e') (p < 0.01), and strain (p = 0.04). Because ATH were younger, adjusted analysis was undertaken in younger HCM patients and compared with ATH with left ventricular wall thickness >13 mm. In this subgroup analysis, the automated model continued to show equal sensitivity, but increased specificity relative to early-to-late diastolic transmitral velocity ratio, e', and strain. CONCLUSIONS: Our results suggested that machine-learning algorithms can assist in the discrimination of physiological versus pathological patterns of hypertrophic remodeling. This effort represents a step toward the development of a real-time, machine-learning-based system for automated interpretation of echocardiographic images, which may help novice readers with limited experience.

Simplifying proximal isovelocity surface area as an assessment method of mitral valve area in patients with rheumatic mitral stenosis by fixing aliasing velocity and mitral valve angle.

UNLABELLED: We aimed to test the ability of a simple equation using proximal isovelocity surface area method (PISA), created by fixing the angle to 100° and the aliasing velocity to 33 cm/s, to calculate mitral valve area (MVA) and assess severity in patients with rheumatic mitral stenosis (MS). METHODS AND RESULTS: In a series of 51 consecutive patients with rheumatic MS, MVA was assessed by four methods, conventional PISA equation (PISAconventional), simple PISA equation (PISAsimple), pressure half time (PHT), and planimetry (PLN) which was taken as the reference method. All methods correlated significantly with PLN with the highest correlation found in case of PISAconventional and PISAsimple (r = 0.97, 0.96, p < 0.001), while the correlation in case PHT was relatively weaker (r = 0.69, p < 0.001). Bland-Altman analysis revealed that the level of agreement with PLN was better in case of both PISA methods than PHT and, moreover, were close to each other. The number of cases that showed agreement of severity grade with planinetry was better in case of PISAconventional (42 cases) and PISAsimple (44 cases) than that in case of PHT (34 cases, p = 0.037). Finally, the measure of agreement with Cohen's Kappa test was better in case of PISAconventional and PISAsimple than that in case of PHT. CONCLUSION: Provided that aliasing velocity is fixed at 33 cm/s, PISA can effectively predict mitral valve area and severity of MS by a simple equation, with the advantage of easy and accurate calculation over other methods.

Tissue Doppler imaging-derived myocardial acceleration during isovolumetric contraction predicts pulmonary capillary wedge pressure in patients with reduced ejection fraction.

BACKGROUND: Tissue Doppler imaging-obtained isovolumetric myocardial acceleration (IVA) is load independent, reportedly predicts systolic functions, and correlates with exercise capacity in patients with reduced ejection fraction (EF). We hypothesized that IVA correlates with the pulmonary capillary wedge pressure (PCWP) in patients with reduced EF. METHODS AND RESULTS: Of 113 patients, correlations between PCWP and IVA were done for all patients, 48 patients with EF ≥55%, and 65 patients with EF <55%. Results were compared to the correlation between PCWP and other echocardiographic predictors. IVA correlated moderately with PCWP in all patients (r=0.54, P<0.0001) and was comparable to the E/A and E/e' ratios. In patients with EF ≥55%, IVA lost correlation and the only predictor was the E/e' ratio (r=0.08, 0.58, P=0.58, <0.0001). In patients with EF <55%, IVA was better than E/A and E/e' (r=0.72, 0.61, 0.51, P<0.0001), especially for atrial fibrillation or when E/e' fell between 8 and 15. Furthermore, IVA >1.60 m/s(2) can predict PCWP ≥15 mmHg, with a sensitivity of 95%, specificity of 73%, and an area under the curve of 0.867 (P<0.0001). CONCLUSIONS: IVA can predict PCWP in patients with reduced EF, and can be considered an alternative to the E/e' ratio for patients with atrial fibrillation or E/e' ratio between 8 and 15.

Comparison of mitral valve area by pressure half-time and proximal isovelocity surface area method in patients with mitral stenosis: effect of net atrioventricular compliance.

AIMS: The aim of this study was to test the hypothesis that, unlike calculation of the mitral valve area (MVA) with the pressure half-time method (PHT), the proximal isovelocity surface area method (PISA) is not affected by changes in net atrioventricular compliance (C(n)). METHODS AND RESULTS: We studied 51 patients with mitral stenosis (MS) from two centres. MVA was assessed with the PISA (MVA(PISA)), PHT (MVA(PHT)), and planimetry (MVA(PLN), serving as the gold standard) method. C(n) was calculated with a previously validated equation using 2D echocardiography. MVA(PISA) closely correlated with MVA(PLN) (r = 0.96, P < 0.0001), while MVA(PHT) and MVA(PLN) showed a weaker but still good correlation (r = 0.69, P < 0.0001). The correlation between MVA(PHT) and MVA(PLN) for patients with C(n) between 4 and 6 mL/mmHg (considered to be normal) was excellent (r = 0.93, P < 0.0001), but that for patients with C(n) of less than 4 or more than 6 mL/mmHg was not as good (r = 0.64, P < 0.0001). Importantly, a significant inverse correlation was detected between the percentage difference among MVA(PHT), MVA(PLN), and C(n) (r = -0.77, P < 0.0001), but the line of fit was nearly flat for the percentage difference among MVA(PISA), MVA(PLN), and C(n) (r = 0.1, P = 0.388). CONCLUSION: MVA calculated with both the PISA and PHT methods correlated well with MVA calculated with the planimetry method. However, the PISA rather than PHT is recommended for patients with MS and extreme C(n) values because PISA, unlike PHT, is not affected by changes in C(n).