Training and clinical testing of artificial intelligence derived right atrial cardiovascular magnetic resonance measurements.

BACKGROUND: Right atrial (RA) area predicts mortality in patients with pulmonary hypertension, and is recommended by the European Society of Cardiology/European Respiratory Society pulmonary hypertension guidelines. The advent of deep learning may allow more reliable measurement of RA areas to improve clinical assessments. The aim of this study was to automate cardiovascular magnetic resonance (CMR) RA area measurements and evaluate the clinical utility by assessing repeatability, correlation with invasive haemodynamics and prognostic value. METHODS: A deep learning RA area CMR contouring model was trained in a multicentre cohort of 365 patients with pulmonary hypertension, left ventricular pathology and healthy subjects. Inter-study repeatability (intraclass correlation coefficient (ICC)) and agreement of contours (DICE similarity coefficient (DSC)) were assessed in a prospective cohort (n = 36). Clinical testing and mortality prediction was performed in n = 400 patients that were not used in the training nor prospective cohort, and the correlation of automatic and manual RA measurements with invasive haemodynamics assessed in n = 212/400. Radiologist quality control (QC) was performed in the ASPIRE registry, n = 3795 patients. The primary QC observer evaluated all the segmentations and recorded them as satisfactory, suboptimal or failure. A second QC observer analysed a random subcohort to assess QC agreement (n = 1018). RESULTS: All deep learning RA measurements showed higher interstudy repeatability (ICC 0.91 to 0.95) compared to manual RA measurements (1st observer ICC 0.82 to 0.88, 2nd observer ICC 0.88 to 0.91). DSC showed high agreement comparing automatic artificial intelligence and manual CMR readers. Maximal RA area mean and standard deviation (SD) DSC metric for observer 1 vs observer 2, automatic measurements vs observer 1 and automatic measurements vs observer 2 is 92.4 ± 3.5 cm2, 91.2 ± 4.5 cm2 and 93.2 ± 3.2 cm2, respectively. Minimal RA area mean and SD DSC metric for observer 1 vs observer 2, automatic measurements vs observer 1 and automatic measurements vs observer 2 was 89.8 ± 3.9 cm2, 87.0 ± 5.8 cm2 and 91.8 ± 4.8 cm2. Automatic RA area measurements all showed moderate correlation with invasive parameters (r = 0.45 to 0.66), manual (r = 0.36 to 0.57). Maximal RA area could accurately predict elevated mean RA pressure low and high-risk thresholds (area under the receiver operating characteristic curve artificial intelligence = 0.82/0.87 vs manual = 0.78/0.83), and predicted mortality similar to manual measurements, both p < 0.01. In the QC evaluation, artificial intelligence segmentations were suboptimal at 108/3795 and a low failure rate of 16/3795. In a subcohort (n = 1018), agreement by two QC observers was excellent, kappa 0.84. CONCLUSION: Automatic artificial intelligence CMR derived RA size and function are accurate, have excellent repeatability, moderate associations with invasive haemodynamics and predict mortality.

Right ventricular size and function under riociguat in pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension (the RIVER study).

BACKGROUND: Riociguat is a soluble guanylate cyclase stimulator approved for pulmonary arterial hypertension (PAH) and chronic thromboembolic pulmonary hypertension (CTPEH). The objective of this study was to evaluate right heart size and function assessed by echocardiography during long term treatment with riociguat. METHODS: Patients who started riociguat treatment (1.0-2.5 mg tid) within the trials phase II, PATENT, PATENTplus, EAS, CHEST and continued treatment for 3-12 months were included in this study. Echocardiography was analysed off-line at baseline, after 3, 6 and 12 months by investigators who were blinded to clinical data. Last and baseline observation carried forward method (LOCF, BOCF) were performed as sensitivity analysis. RESULTS: Seventy-one patients (45% PAH, 55% CTEPH; 53.5% female; 60 ± 13 years, mean pulmonary arterial pressure 46 ± 10 mmHg, mean PVR 700 ± 282dynes·sec·cm-5) were included. After 6 months, RA and RV area, RV thickness tricuspid regurgitation velocity showed a significant reduction. After 12 months, patients receiving riociguat therapy showed a significant reduction in right atrial (- 2.6 ± 4.4 cm2, 95% CI -3.84, - 1.33; p < 0.001, n = 49) and right ventricular (RV) area (- 3.5 ± 5.2 cm2, 95% CI -5.1, - 1.9; p < 0.001; n = 44), RV thickness (- 0.76 ± 2.2 mm, 95% CI -1.55, 0.03; n = 32), and a significant increase in TAPSE (2.95 ± 4.78 mm, 95% CI 1.52, 4.39; n = 45) and RV fractional area change (8.12 ± 8.87 mm, 95% CI 4.61, 11.62; n = 27). Both LOCF and BOCF showed similar results but lower effect sizes. CONCLUSION: Patients under long-term treatment with riociguat show significantly reduced right heart size and improved RV function in PAH and CTEPH. Further controlled prospective studies are needed to confirm these results.

Validation of Right Atrial Area as a Measure of Right Atrial Size and Normal Values of in Healthy Pediatric Population by Two-Dimensional Echocardiography.

Right atrial (RA) size is a prognostic indicator for heart failure and cardiovascular death in adults. Data regarding use of RA area (RAA) by two-dimensional echocardiography as a surrogate for RA size and allometric modeling to define appropriate indexing of the RAA are lacking. Our objective was to validate RAA as a reliable measure of RA size and to define normal reference values by transthoracic echocardiography (TTE) in a large population of healthy children and develop Z-scores using a validated allometric model for indexing RAA independent of age, sex, and body size. Agreement between RAA and volume by 2D, 3D TTE, and MRI was assessed. RAA not volume by 2D TTE is an excellent surrogate for RA size. RAA/BSA1 has an inverse correlation with BSA with a residual relationship to BSA (r = - 0.54, p < 0.0001). The allometric exponent (AE) derived for the entire cohort (0.85) also fails to eliminate the residual relationship. The entire cohort divided into two groups with a BSA cut-off of 1 m2 to provide the best-fit allometric model (r = 0). The AE by least square regression analysis for each group is 0.95 and 0.88 for BSA < 1 m2 and > 1 m2, respectively, and was validated against an independent sample. The mean indexed RAA ± SD for BSA ≤ 1 m2 and > 1 m2 is 9.7 ± 1.3 cm2 and 8.7 ± 1.3 cm2, respectively, and was used to derive Z-scores. RAA by 2D TTE is superior to 2D or 3D echocardiography-derived RA volume as a measure of RA size using CMR as the reference standard. RAA when indexed to BSA1, decreases as body size increases. The best-fit allometric modeling is used to create Z scores. RAA/BSA0.95 for BSA < 1 m2 and RAA/BSA0.88 for those with BSA > 1 m2 can be used to derive Z scores.

Diastolic Dysfunction Is a Contributing Factor to Exercise Intolerance in COPD.

Right ventricular (RV) systolic failure is rare in patients with COPD, but they often develop RV diastolic dysfunction. Left ventricular (LV) diastolic dysfunction is also common in this population. Nevertheless, data are scarce regarding the effect of diastolic dysfunction on the functional capacity in patients with COPD. We investigated the correlation between echocardiographic parameters of RV and LV diastolic function and the exercise capacity in COPD, by using conventional echocardiographic methods and tissue Doppler imaging. 65 patients with COPD (61 ± 9 years) in stages GOLD II-IV were investigated. Functional capacity was measured with 6-minute walk test (6MWT). Right (RA) and left atrial (LA) area index were measured; collapsibility index inferior vena cava was calculated. Parameters of the mitral and tricuspid inflow (E, A) as well as annular systolic (S), early- (e') and late- (a') diastolic myocardial longitudinal velocities were measured. E/A, E/e' and e'/a' ratios were calculated. 6MWT distance was 330 ± 76 m. LV diastolic dysfunction was found in 48 (74%) patients. LV and RV filling pressures were elevated in 28 (43%) and in 29 (45%) patients, respectively. In the left heart, LA area index showed significant correlation with the functional capacity (r = -0.319; p = 0.011). In stepwise multiple linear regression analysis tricuspid e'/a' (r = 0.611; p = 0.000), collapsibility index (r = 0.505; p = 0.000), RA area index (r = -0.445; p = 0.000) and body surface area (r = 0.314; p = 0.011) were independent predictors of 6MWT distance. Right ventricular diastolic function and filling pressure have strong influence on the functional capacity in patients with COPD.

Imaging and Risk Stratification in Pulmonary Arterial Hypertension: Time to Include Right Ventricular Assessment.

Background: Current European Society of Cardiology and European Respiratory Society guidelines recommend regular risk stratification with an aim of treating patients with pulmonary arterial hypertension (PAH) to improve or maintain low-risk status (<5% 1-year mortality). Methods: Consecutive patients with PAH who underwent cardiac magnetic resonance imaging (cMRI) were identified from the Assessing the Spectrum of Pulmonary hypertension Identified at a Referral centre (ASPIRE) registry. Kaplan-Meier survival curves, locally weighted scatterplot smoothing regression and multi-variable logistic regression analysis were performed. Results: In 311 consecutive, treatment-naïve patients with PAH undergoing cMRI including 121 undergoing follow-up cMRI, measures of right ventricular (RV) function including right ventricular ejection fraction (RVEF) and RV end systolic volume and right atrial (RA) area had prognostic value. However, only RV metrics were able to identify a low-risk status. Age (p < 0.01) and RVEF (p < 0.01) but not RA area were independent predictors of 1-year mortality. Conclusion: This study highlights the need for guidelines to include measures of RV function rather than RA area alone to aid the risk stratification of patients with PAH.

Right Ventricular Diastolic Performance in Patients With Chronic Thromboembolic Pulmonary Hypertension Assessed by Echocardiography.

Background: There have been no systemic studies about right heart filling pressure and right ventricular (RV) distensibility in patients with chronic thromboembolic pulmonary hypertension (CTEPH). Therefore, we aimed to explore combinations of echocardiographic indices to assess the stages of RV diastolic dysfunction. Methods and Results: We recruited 32 healthy volunteers and 71 patients with CTEPH. All participants underwent echocardiography, cardiac catheterization (in patients with CTEPH), and a 6-min walk test (6MWT). The right atrial (RA) end-systolic area was adjusted for body surface area (BSA) (indexed RA area). RV global longitudinal diastolic strain rates (SRs) and RV ejection fraction (EF) were measured by speckle tracking and three-dimensional echocardiography (3D echo), respectively. All 71 patients with CTEPH underwent pulmonary endarterectomy. Of the 71 patients, 52 (73%) had decreased RV systolic function; 12 (16.9%), 26 (36.6%), and 33 (46.5%) patients had normal RV diastolic pattern, abnormal relaxation (stage 1), and pseudo-normal patterns (stage 2), respectively. The receiver operating characteristic curve analysis showed that the optimal cut-off values of early diastolic SR <0.8 s-1 and indexed RA area > 8.8 cm2/BSA had the best accuracy in identifying patients with RV diastolic dysfunction, with 87% sensitivity and 82% specificity. During a mean follow-up of 25.2 months after pulmonary endarterectomy, the preoperative indexed RA area was shown as an independent risk factor of the decreased 6MWT distance. Conclusions: Measuring early diastolic SR and indexed RA area would be useful in stratifying RV diastolic function.

Relationship between severity of pulmonary hypertension and coronary sinus diameter.

INTRODUCTION AND OBJECTIVE: We investigated the relationship between coronary sinus (CS) diameter and pulmonary artery systolic pressure (PASP) in patients with pulmonary hypertension (PH) and normal left ventricular systolic function. METHODS: A total of 155 participants referred for transthoracic echocardiography were included in the study. The study population consisted of 100 patients with chronic PH and 55 control subjects. Patients with PH were divided into two groups according to PASP: those with PASP 36-45 mmHg, the mild PH group (n=53); and those with PASP >45 mmHg, the moderate to severe PH group (n=47). CS diameter was measured from the posterior atrioventricular groove in apical 4-chamber view during ventricular systole according to the formula: mean CS=(proximal CS+mid CS+distal CS)/3. RESULTS: Mean CS diameter was significantly higher in the moderate to severe PH group than in the controls and in the mild PH group (1.12±0.2 cm vs. 0.82±0.1 cm and 0.87±0.1 cm, respectively; p<0.001). It was significantly correlated with right atrial (RA) area (r=0.674, p<0.001), RA pressure (r=0.458, p<0.001), PASP (r=0.562, p<0.001), inferior vena cava diameter (r=0.416, p<0.001), right ventricular E/A ratio (r=-0.290, p<0.001), and E/Em ratio (r=0.235, p=0.004). RA area (ß=0.475, p<0.001) and PASP (ß=0.360, p=0.002) were found to be independent predictors of CS diameter. CONCLUSIONS: A dilated CS was associated with moderate to severe pulmonary hypertension, and RA area and PASP were independent predictors of CS diameter.

Proximal isovelocity surface area by single-beat three-dimensional color Doppler echocardiography applied for tricuspid regurgitation quantification.

BACKGROUND: The two-dimensional (2D) proximal isovelocity surface area (PISA) method has known technical limitations, mainly the geometric assumptions of PISA shape required to calculate effective regurgitant orifice area (EROA). Recently developed single-beat real-time three-dimensional (3D) color Doppler imaging allows the direct measurement of PISA without geometric assumptions and has already been validated for mitral regurgitation assessment. The aim of this study was to apply this novel method in patients with chronic tricuspid regurgitation (TR). METHODS: Ninety patients with chronic TR were enrolled. EROA and regurgitant volume (Rvol) were assessed using transthoracic 2D and 3D PISA methods. Quantitative Doppler and 3D transthoracic planimetry of EROA were used as reference methods. RESULTS: Both EROA and Rvol assessed using the 3D PISA method had better correlations with the reference methods than using conventional 2D PISA, particularly in the assessment of eccentric jets. On the basis of 3D planimetry-derived EROA, 35 patients had severe TR (EROA ≥ 0.4 cm(2)). Among these 35 patients, 25.7% (n = 9) were underestimated as having nonsevere TR (EROA ≤ 0.4 cm(2)) using the 2D PISA method. In contrast, the 3D PISA method had 94.3% agreement (33 of 35) with 3D planimetry in classifying severe TR. Good intraobserver and interobserver agreement for 3D PISA measurements was observed, with intraclass correlation coefficients of 0.92 and 0.88 respectively. CONCLUSIONS: TR quantification using PISA by single-beat real-time 3D color Doppler echocardiography is feasible in the clinical setting and more accurate than the conventional 2D PISA method.

Three-dimensional color Doppler echocardiographic quantification of tricuspid regurgitation orifice area: comparison with conventional two-dimensional measures.

BACKGROUND: Three-dimensional (3D) color Doppler echocardiography (CDE) provides directly measured vena contracta area (VCA). However, a large comprehensive 3D color Doppler echocardiographic study with sufficiently severe tricuspid regurgitation (TR) to verify its value in determining TR severity in comparison with conventional quantitative and semiquantitative two-dimensional (2D) parameters has not been previously conducted. The aim of this study was to examine the utility and feasibility of directly measured VCA by 3D transthoracic CDE, its correlation with 2D echocardiographic measurements of TR, and its ability to determine severe TR. METHODS: Ninety-two patients with mild or greater TR prospectively underwent 2D and 3D transthoracic echocardiography. Two-dimensional evaluation of TR severity included the ratio of jet area to right atrial area, vena contracta width, and quantification of effective regurgitant orifice area using the flow convergence method. Full-volume breath-hold 3D color data sets of TR were obtained using a real-time 3D echocardiography system. VCA was directly measured by 3D-guided direct planimetry of the color jet. Subgroup analysis included the presence of a pacemaker, eccentricity of the TR jet, ellipticity of the orifice shape, underlying TR mechanism, and baseline rhythm. RESULTS: Three-dimensional VCA correlated well with effective regurgitant orifice area (r = 0.62, P < .0001), moderately with vena contracta width (r = 0.42, P < .0001), and weakly with jet area/right atrial area ratio. Subgroup analysis comparing 3D VCA with 2D effective regurgitant orifice area demonstrated excellent correlation for organic TR (r = 0.86, P < .0001), regular rhythm (r = 0.78, P < .0001), and circular orifice (r = 0.72, P < .0001) but poor correlation in atrial fibrillation rhythm (r = 0.23, P = .0033). Receiver operating characteristic curve analysis for 3D VCA demonstrated good accuracy for severe TR determination. CONCLUSIONS: Three-dimensional VCA measurement is feasible and obtainable in the majority of patients with mild or greater TR. Three-dimensional VCA measurement is also feasible in patients with atrial fibrillation but performed poorly even with <20% cycle length variation. Three-dimensional VCA has good cutoff accuracy in determining severe TR. This simple, straightforward 3D color Doppler measurement shows promise as an alternative for the quantification of TR.