Search

Background Left ventricular outflow tract diameter (LVOTd) is routinely measured to calculate stroke volume and estimate aortic valve area by the continuity equation. Despite LVOTd being regularly measured clinically, significant inter- and intraobserver variability is evident. This variability is highly impactful on both aortic stenosis evaluation and cardiac output calculation due to the squaring of the LVOT radius. Purpose We aimed to investigate if LVOTd measurements from clinical echocardiographic examinations could be used in a deep learning (DL) model to automatically perform LVOTd measurements with equivalent accuracy and improved consistency compared to current practice. Methods Data was collected from clinical echocardiographic examinations performed on 656 consecutive patients admitted to the cardiac catheterization laboratory at a university hospital in January – December 2018. Parasternal views with cardiologist annotated LVOTd coordinates were assessed for 1314 echocardiographic still images. The quality of the still image and annotated LVOT ground truth were individually graded as high, medium and low by experienced cardiologists to establish a rigorous training basis. Spatial geometry data was preserved for each still image in order to distinguish between different degrees of image zoom. Data was randomly split into training, validation and testing sets (68%, 17%, 15%). A fully convolutional network based on the Resnet50 architecture was used with a custom loss function with heatmap regression. Image augmentations were added to extend the dataset. Results When including echocardiographic images of any quality (n=1314) in the model training and inference, the median absolute difference between cardiologist LVOTd and DL LVOTd was 0.97 mm (95% Confidence interval (CI) 0.79–1.14). Using only high and medium quality still images and ground truth (n=869) in the training and inference, median absolute difference decreased to 0.81 mm (95% CI 0.60–0.96). Adding image augmentations to this dataset further improved the model, resulting in a median LVOTd absolute difference of 0.66 mm (95% CI 0.51–0.78). The LVOTd error in inference increased with decreasing image quality, as shown in Figure 1, with two predictions (0.9%) failing completely. Conclusion Deep learning models are capable of measuring LVOTd with comparable accuracy to cardiologists when trained on clinical data. Data quality affects both training and inference. Even with a slightly lower accuracy when used on lower quality echocardiographic images, DL-assisted LVOT measurement has a clear potential to increase repeatability and consistency of LVOTd measurements. Funding Acknowledgement Type of funding sources: Public Institution(s). Main funding source(s): Research Council of Norway (Norges forskningsråd) Figure 1

To summarize literature on the application of myocardial strain in patients with heart failure published since 2018.Left ventricular (LV) and right ventricular strain provides important prognostic information in patients with acute and chronic heart failure and new insights in disease mechanisms in amyloidosis. The addition of left atrial strain to current societal diastolic dysfunction criteria may improve detection of clinically relevant diastolic dysfunction. The recently developed method for noninvasive estimation of myocardial work incorporates loading conditions into the evaluation of LV performance that may be important for selection of patients for cardiac resynchronization therapy.Evidence linking myocardial strain to adverse outcomes in heart failure is steadily being developed. Although GLS seems to be ready for clinical use, further validation and standardization of RV, LA strain and myocardial work is needed.

Myocardial deformation or strain is the universal property of contracting cardiac muscle. Deformation is defined in physics as relative change of length (and is therefore unitless and usually given as percentage) and in cardiac imaging it is thus algebraically negative for shortening or positive for thickening. There are several definitions of strain—Lagrangian strain refers to a fixed baseline distance and Eulerian (or natural) strain—to a dynamically changing reference length, representing a time integral of strain rate (which can be obtained by tissue Doppler). Measurements of strains are usually obtained by greyscale image quantification modality—speckle-tracking echocardiography (STE) which analyses myocardial motion by tracking and matching naturally occurring markers of myocardial texture, described as speckles. Echocardiographic speckles represent interference pattern of subtle myocardial scatters and can be followed from frame to frame by dedicated software to define the displacement of the myocardium within the interval between consecutive frames (inverse of frame rate).

Background Recent evidence suggests that an implantable cardioverter defibrillator (ICD) in non-ischemic cardiomyopathy (NICM) may not offer mortality benefit. We aimed to investigate if etiology of heart failure and strain echocardiography can improve risk stratification of life threatening ventricular arrhythmia (VA) in heart failure patients. Methods This prospective multi-center follow-up study consecutively included NICM and ischemic cardiomyopathy (ICM) patients with left ventricular ejection fraction (LVEF) 70 ms had highest VA incidence with an event rate of 16%/year. Conclusion Patients with NICM and normal MD had low arrhythmic event rate, comparable to the general population. Patients with ICM and MD >70 ms had the highest risk of VA. Combining heart failure etiology and strain echocardiography may classify heart failure patients in low, intermediate and high risk of VA and thereby aid ICD decision strategies.

Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Research Council of Norway Background Speckle tracking echocardiography (STE) provides quantification of left ventricular (LV) deformation and is useful in the assessment of LV function. STE is increasingly being used clinically, and every effort to simplify and standardise STE is important. Manual outlining of regions of interest (ROIs) is labour intensive and may influence assessment of strain values. Purpose We hypothesised that ROIs from clinical echocardiographic exams could be used to train a deep learning (DL) model which could automate strain calculation by predicting ROIs with comparable fidelity to trained cardiologists. Methods Data consisted of still frames with cardiologist-defined ROIs from 435 clinical echocardiographic exams from a university hospital outpatient clinic. Exams included healthy subjects and patients with ischemic heart disease, heart failure, valvular disease and conduction abnormalities. Three frames surrounding mid-systole from standard apical views were paired with the mid-systolic ROI for training. Image quality was classified as high, medium or low by an experienced cardiologist prior to model training, and low quality images were excluded. The dataset was randomly split into 68%/17%/15% sets for training, validation and testing. A standard Unet architecture was used, with a combination of binary cross entropy and Dice loss functions. A total of 1025 images were used for final model training and testing, with augmentation used to extend the dataset. Predicted ROIs were reintroduced into commercially available echocardiographic analysis software to automatically calculate longitudinal strain (LS) values. Results The average absolute difference between manually measured and DL-assisted LS was 0.98 percentage points (95% confidence interval 0.82-1.13) for the final model. A Bland-Altman plot revealed very few outliers and no bias in the difference between manual and DL-assisted strain measurements. Conclusion The current study demonstrates that DL-assisted, automated strain measurement is feasible, and provides strain results within an acceptable range of intra- and interobserver variation. Employing DL in echocardiographic analysis could further facilitate adoption of STE parameters in clinical practice and research, and improve reproducibility. Abstract Figure. Bland-Altman plot. Manual and DL ROIs

Background: In patients with aortic stenosis, subtle alterations in myocardial mechanics can be detected by speckle-tracking echocardiography before reduction of left ventricular ejection fraction (LVEF). Methods: In this prospective study, 162 patients with aortic stenosis with an average aortic valve area of 0.7 ± 0.2 cm2 and a mean LVEF of 60 ± 11% were included. Global longitudinal strain (GLS) and mechanical dispersion (SD of time from Q/R on the electrocardiogram to peak strain in 16 left ventricular segments) were assessed using echocardiography, and all-cause mortality (n = 37) was recorded during 37 ± 13 months of follow-up. Results: Overall, nonsurvivors had more pronounced mechanical dispersion and worse GLS compared with survivors (74 ± 24 vs 61 ± 18 msec [P < .01] and −14.5 ± 4.4% vs −16.7 ± 3.6% [P < .01], respectively). In the 42 conservatively treated patients without surgical aortic valve replacement, a similar pattern was observed in nonsurvivors versus survivors (mechanical dispersion, 80 ± 24 vs 57 ± 14 msec [P < .01]; GLS, −14.0 ± 4.9% vs −17.1 ± 3.8% [P = .04], respectively). Mechanical dispersion was significantly associated with mortality (hazard ratio per 10-msec increase, 1.23; 95% CI, 1.07–1.42; P < .01) in a Cox model adjusted for LVEF and with aortic valve replacement treatment as a time-dependent covariate. Continuous net reclassification improvement showed that mechanical dispersion was incremental to LVEF, GLS, and valvulo-arterial impedance when adjusting for aortic valve replacement treatment in the total population. Conclusion: Increased mechanical dispersion may be a risk marker providing novel prognostic information in patients with aortic stenosis.

ObjectivesThe physiological determinants of left ventricular (LV) mechanical dispersion (MD) are not fully explored. We aimed to investigate the impact of afterload reduction and changes in ventricular conduction on LV MD after transcatheter aortic valve implantation (TAVI).MethodsPatients with severe aortic stenosis (AS) were examined in a prospective, repeated measures observational cohort study before and after an uncomplicated transfemoral TAVI in a single tertiary centre. LV MD was assessed by speckle tracking echocardiography. Valvulo-arterial impedance (ZVA) was used as a measure of global afterload.ResultsWe included 140 consecutive patients (83±8 years old, 49% women, logistic EuroSCORE 16±10) with severe AS (valve area 0.7±0.2 cm2, mean transvalvular gradient 54±18 mm Hg) and a relatively preserved LV ejection fraction (52%±11%). After TAVI, we observed favourable changes in transvalvular gradients and ZVA in all patients. Compared with baseline, postprocedural MD was significantly lower in 108 patients with unchanged ventricular conduction (55±17 ms vs 51±17 ms, p=0.02) and higher in 28 patients with TAVI-induced left bundle branch block (51±13 ms vs 62±19 ms, p≤0.001). During 22±9 months observation, 22 patients died. Postprocedural MD was associated with mortality in a univariate Cox regression model (HR=1.24 (1.01–1.52), pConclusionsIsolated afterload reduction was associated with reduction of MD, while concomitant impairment of ventricular conduction resulted in a more pronounced MD after TAVI, indicating that loading conditions and conduction should be considered when evaluating MD. A pronounced postprocedural LV MD was associated with mortality.

Background Left bundle branch block (LBBB) frequently complicates transcatheter aortic valve implantation (TAVI), but may not induce classical mechanical dyssynchrony typically observed in heart failure patients with LBBB. Non-invasive estimated myocardial work by pressure strain-loops is a novel method for evaluation of left ventricular (LV) performance including loading conditions and identifies responders to cardiac resynchronization therapy. Assessment of myocardial work in TAVI-induced LBBB has, however, not been performed earlier. Purpose We aimed to evaluate myocardial performance in TAVI-induced LBBB by myocardial work. Methods Patients with severe aortic stenosis (AS) and normal ventricular conduction undergoing transfemoral TAVI were included. TAVI-induced LBBB was defined according to strict ECG criteria. In addition to standard echocardiographic parameters, we performed speckle tracking echocardiography (STE) within 2 days post procedure. Post-procedural myocardial work was estimated by commercially available software, combining LV strain by STE and a non-invasively estimated LV pressure curve. Peak systolic LV pressure was assumed to be equal to the sum of peak arterial cuff pressure and mean aortic valve pressure gradient. The proportion of constructive work relative to the sum of wasted and constructive work was expressed as myocardial work efficiency (Figure). A global work efficiency < 90 % was considered abnormal according to the recently published EACVI NORRE study. Results We included 109 consecutive patients undergoing TAVI (83 ± 7 years old, 54 % women, logistic Euroscore 16 ± 10) with severe AS (valve area 0.7 ± 0.2 cm2, mean transvalvular gradient 54 ± 18 mmHg), QRS duration 94 ± 10 ms, and a relatively preserved LV ejection fraction (54 ± 9%). New-onset LBBB was observed in 28 patients (26 %) after TAVI. Although post-procedural myocardial work efficiency was slightly better in patients who retained normal ventricular conduction than those with TAVI-induced LBBB (93.6 ± 2.5% vs 91.2 ± 3.4%, p = 0.004), both groups of patients had global myocardial work efficiency within the normal reference range. Conclusions Patients with TAVI-induced LBBB have preserved work efficiency despite gross electrical block satisfying strict ECG criteria. The reduced myocardial work efficiency in TAVI-induced LBBB was within the reference range of normal in the recently published EACVI NORRE study. The preservation of myocardial work in TAVI-induced LBBB may explain why LBBB as a complication to TAVI does not necessarily imply poor prognosis. Abstract P1575 Figure.

Background Secretoneurin (SN) concentrations provide important prognostic information in patients with myocardial dysfunction. Whether preoperative SN concentrations improve risk assessment in patients with moderate to severe aortic stenosis (AS) is unknown. Methods We included 57 patients with moderate to severe AS referred for presurgical evaluation. All patients were examined with comprehensive echocardiography, electrocardiogram (ECG), and biochemical measurements and compared to 10 age- and sex-matched healthy subjects. Results Median (quartile 1–3) SN concentrations were 141 (121–163) pmol/L in AS patients and 132 (106–148) pmol/L in control subjects (p = .17). Lower estimated creatinine clearance and use of diuretics, but not standard ECG or echocardiographic indices and cardiac biomarkers, were associated with increasing SN concentrations. Fifteen patients (26%) died during 3.5 years median follow-up. SN concentrations were higher in non-survivors than survivors: 156 (133–209) vs. 140 (116–155) pmol/L, p = .007. Higher SN concentrations were associated with increased risk of mortality also after adjustment for established risk indices, biomarkers, and status regarding valvular surgery: hazard ratio per lnSN 15.13 (95% CI 1.05–219.00); p = .046. Receiver operating characteristics area under the curve for SN to predict mortality was 0.74 (95% CI 0.60–0.88) compared to 0.73 (0.59–0.87) for high-sensitivity cardiac troponin T and 0.67 (0.51–0.82) for N-terminal pro-B-type natriuretic peptide. The previously identified cut-off of SN >204 pmol/L in cardiac surgical patients predicted mortality also in this cohort. Conclusions SN concentrations improve risk assessment in patients with moderate to severe AS by providing additional prognostic information to established risk indices such as echocardiography, ECG, and established cardiac biomarkers.

Aims Left bundle branch block (LBBB) is a frequent conduction abnormality after transcatheter aortic valve implantation (TAVI). We aimed to investigate how TAVI procedure related conduction abnormalities influence ventricular mechanics and prognosis, with particular focus on new-onset persistent LBBB. Methods and results A total of 140 consecutive patients with severe aortic stenosis (83 ± 8 years old, 49% women) undergoing TAVI in a single tertiary centre were included in a repeated measures study. Changes in myocardial function and contraction patterns were investigated in relation to changes in electrical conduction and afterload by speckle tracking echocardiography. Whether patients with new-onset LBBB acquired classical dyssynchronous contractions was assessed by longitudinal strain in apical four-chamber view. Global longitudinal strain improvement was seen in all patients (−15.1 ± 4.3 vs. −16.1 ± 3.9%, P

Left ventricular (LV) ejection fraction (LVEF) is the most validated and commonly used echocardiographic measure of systolic function. LVEF has a unique position in cardiology having severed as selection criteria for therapeutic trials that constitute the evidence base of today's treatment recommendations. Assessment of LV systolic function by global longitudinal strain (GLS) from speckle tracking echocardiography (STE) is a sensitive and feasible method that overcomes many of the limitations of LVEF, including reproducibility issues of serial testing and detection of LV dysfunction in pathologically remodeled hearts. This review discusses the role of STE as a complementary method to LVEF in estimation of LV systolic function.

Aims We aimed to assess the relation between number of pregnancies and cardiac structure, function, and arrhythmic events in women with arrhythmogenic cardiomyopathy (AC). Methods and results We included female AC patients in a cross-sectional study. Number of pregnancies and pregnancy related symptoms were recorded. Ventricular arrhythmias were defined as aborted cardiac arrest, sustained ventricular tachycardia, or appropriate implantable cardioverter-defibrillator therapy. Right and left ventricular dimensions and function, including strain analyses, were assessed by echocardiography and magnetic resonance imaging. We created a new AC severity score to grade the severity of AC disease. We included 77 women (age 47 ± 16, 43 probands and 34 AC mutation positive female relatives), 19 ± 14 years after last pregnancy. Median number of pregnancies was 2 (0–4); 19 had no previous pregnancies, 16 had 1 pregnancy, 30 had 2, and 12 had ≥3 pregnancies. Presence of a definite AC diagnosis (P = 0.36), severity of AC disease (P = 0.53), and arrhythmic events (P = 0.25) did not differ between groups of pregnancies. Number of pregnancies was related to increased right ventricular outflow tract diameter in single variable analyses [odds ratio (OR) 1.76, 95% confidence interval (CI) 1.08–2.87; P = 0.02], but not when adjusted for body surface area and age (OR 1.56, 95% CI 0.91–2.66; P = 0.11). The number of pregnancies was not associated with any other measures of cardiac structure and function. Conclusion Higher number of pregnancies did not seem to relate to a worse phenotype in women with AC.