Search

Left ventricular dilatation (LVD) and left ventricular hypertrophy (LVH) are risk factors for heart failure, and their detection improves heart failure screening. This study aimed to investigate the ability of deep learning to detect LVD and LVH from a 12-lead electrocardiogram (ECG). Using ECG and echocardiographic data, we developed deep learning and machine learning models to detect LVD and LVH. We also examined conventional ECG criteria for the diagnosis of LVH. We calculated the area under the receiver operating characteristic (AUROC) curve, sensitivity, specificity, and accuracy of each model and compared the performance of the models. We analyzed data for 18,954 patients (mean age (standard deviation): 64.2 (16.5) years, men: 56.7%). For the detection of LVD, the value (95% confidence interval) of the AUROC was 0.810 (0.801-0.819) for the deep learning model, and this was significantly higher than that of the logistic regression and random forest methods (P0.001). The AUROCs for the logistic regression and random forest methods (machine learning models) were 0.770 (0.761-0.779) and 0.757 (0.747-0.767), respectively. For the detection of LVH, the AUROC was 0.784 (0.777-0.791) for the deep learning model, and this was significantly higher than that of the logistic regression and random forest methods and conventional ECG criteria (P0.001). The AUROCs for the logistic regression and random forest methods were 0.758 (0.751-0.765) and 0.716 (0.708-0.724), respectively. This study suggests that deep learning is a useful method to detect LVD and LVH from 12-lead ECGs.

Preservation of the mitral valve (MV) size is essential for valve function, and a reduced MV coaptation-zone area increases the risk of developing functional mitral regurgitation (FMR). We aimed to determine if the MV leaflet and coaptation-zone areas were associated with the severity of atherosclerosis assessed by cardio-ankle vascular index (CAVI) in patients with normal left ventricle (LV) systolic function and size by real-time 3D echocardiography (RT3DE).We performed RT3DE analysis in 66 patients with normal LV size and ejection fraction who underwent 2D echocardiography and CAVI. MV coaptation-zone areas were measured by custom 3D software and indexed by body surface area (BSA). The associations of clinical factors and mean CAVI with MV leaflet and coaptation-zone areas were evaluated by univariable and multivariable linear regression analyses.On univariable analysis, MV leaflet area/BSA was significantly associated with age (r = -0.335, P = 0.0069) and mean CAVI (r = -0.464, P < 0.001), and MV coaptation-zone area was significantly associated with age (r = -0.626, P < 0.001), hypertension (r = -0.626, P < 0.001), dyslipidemia (r = -0.626, P < 0.001), E/e' (r = -0.626, P < 0.001), and CAVI (r = -0.740, P < 0.001). On multivariable analysis, mean CAVI was independently associated only with MV leaflet area/BSA (standardized coefficient = -0.611, P < 0.001) and MV coaptation-zone area/BSA (standardized coefficient = -0.74, P < 0.001).In patients with normal LV systolic function and size, MV leaflet and coaptation-zone areas might be reduced according to advancing atherosclerosis. Patients with atherosclerosis might be at increased risk of developing FMR.

Enlargement of the mitral valve (MV) has gained attention as a compensatory mechanism for functional mitral regurgitation (FMR). We aimed to determine if MV leaflet area is associated with MV coaptation-zone area and identify the clinical factors associated with MV leaflet size and coaptation-zone area in patients with normal left ventricle (LV) systolic function and size using real-time 3D echocardiography (RT3DE).We performed RT3DE in 135 patients with normal LV size and ejection fraction. MV leaflet and coaptation-zone areas were measured using custom 3D software. The clinical factors associated with MV leaflet and coaptation-zone areas were evaluated using univariate and multivariate linear regression analyses.There was a significant relationship between MV leaflet and coaptation-zone areas (r = 0.499, P < 0.001). MV leaflet area was strongly associated with body surface area (BSA) (r = 0.905, P < 0.001) rather than LV size and age. MV leaflet area/BSA was independently associated with male gender (P = 0.002), lower diastolic blood pressure (P = 0.042), and LV end-diastolic volume (LVEDV) index (P = 0.048); MV coaptation-zone area/BSA was independently associated with lower LVEDV index (P = 0.01).In patients with normal LV systolic function and size, MV leaflet size has a significant impact on competent MV coaptation. MV leaflet area might be intrinsically determined by body size rather than age and LV size, and the MV leaflet area/BSA is relatively constant. On the other hand, some clinical factors might also influence MV leaflet and coaptation-zone area.

Pulmonary hypertension (PH) is defined as a mean pulmonary artery pressure (PAP) ≥ 25 mmHg at rest as assessed by right heart catheterization (RHC), and Doppler-derived systolic PAP (sPAPECHO) or tricuspid regurgitation pressure gradient (TRPG) is widely used to screen for PH. However, the cutoff value of sPAPECHO or TRPG for detecting a mean PAP ≥ 25 mmHg that was determined invasively has not been well defined.We studied 189 patients who underwent RHC. Echocardiography was performed within 24 hours of invasive evaluation, and sPAPECHO was defined as the TRPG with right atrial pressure estimated on the basis of the current guideline.From the receiver operating characteristic (ROC) curve analysis, the optimal sPAPECHO, and TRPG cutoffs for detecting PH were 41 mmHg (sensitivity, 92%; specificity, 91%; area under the curve = 0.95) and 36 mmHg (sensitivity, 90%; specificity, 93%; area under the curve = 0.95), respectively. The area under the TRPG ROC curve was similar to the area under the sPAPECHO ROC curve.Given that Doppler echocardiography is required to accurately detect PH rather than to accurately estimate systolic PAP, our results provide useful information with regard to screening patients for PH and recommending further investigations on PH.

The prognostic value of the right ventricular (RV) systolic to diastolic duration ratio (S/D ratio) in patients with advanced heart failure is not clear.We enrolled 45 patients with DCM (40 ± 13 years, 33 male) who were admitted to our hospital for evaluation or treatment of heart failure. The RV systolic and diastolic durations were measured using continuous Doppler imaging of tricuspid regurgitation, and the RV S/D ratio was calculated. Cardiac events were defined as cardiac death or left ventricular assist device implantation within the first year. Twenty-eight cardiac events occurred. The RV S/D ratio was significantly higher in the event group than in the event-free group (1.8 ± 0.8 versus 1.2 ± 0.5, P = 0.008). Univariate analysis showed that the RV S/D ratio, plasma brain natriuretic peptide concentration, left atrial volume index, and mitral deceleration time were associated with these events. Receiver operating characteristic curve analysis revealed that the optimal RV S/D cutoff value to predict events was 1.2 (sensitivity 79%, specificity 65%, area under the curve 0.745). Kaplan-Meier analysis indicated a significantly higher event rate in patients with an RV S/D ratio > 1.2 (log-rank test, P = 0.003). The addition of an RV S/D ratio > 1.2 improved the prognostic utility of a model that included conventional variables (P = 0.014).In patients with advanced heart failure with DCM, the RV S/D ratio was higher in patients with events than in those without events. The addition of the RV S/D ratio to conventional parameters may provide better prognostic information.

Emerging concerns regarding heart failure, arrhythmia, and sudden death in patients with muscular dystrophy are of significant clinical importance. On the other hand, little attention has been paid to renal dysfunction because these patients have low serum creatinine levels. Serum cystatin C, unaffected by muscle quantity, is a potentially superior marker for estimating renal function. Here, we present cases with muscular dystrophy in which estimated glomerular filtration rate (GFR) by cystatin C (eGFRcys) provided good agreement with simultaneously measured GFR by inulin renal clearance (differences less than 20%). Sudden death with acute heart failure occurred in a patient with underlying renal dysfunction and elevated BNP. Neurologists and cardiologists should evaluate renal function using GFR with cystatin C in patients with muscular dystrophy.