Cardiac remodeling is recognized as an important aspect of cardiovascular disease (CVD) progression. Machine learning (ML) techniques were applied to basic clinical parameters and electrocardiographic features, in order to detect abnormal left ventricular geometry (LVG) even before the onset of left ventricular hypertrophy (LVH), in a population without established CVD. The authors enrolled 528 patients with and without essential hypertension, but no other indications of CVD. All patients underwent a full echocardiographic evaluation and were classified into 3 groups; normal geometry (NG), concentric remodeling without LVH (CR), and LVH. Abnormal LVG was identified as increased relative wall thickness (RWT) and/or left ventricular mass index (LVMi). The authors trained supervised ML models to classify patients with abnormal LVG and calculated SHAP values to perform feature importance and interaction analysis. Hypertension, age, body mass index over the Sokolow‐Lyon voltage, QRS‐T angle, and QTc duration were some of the most important features. Our model was able to distinguish NG from CR+LVH combined, with 87% accuracy on an unseen test set, 75% specificity, 97% sensitivity, and area under the receiver operating curve (AUC/ROC) equal to 0.91. The authors also trained our model to classify NG and CR (NG + CR) against those with LVH, with 89% test set accuracy, 93% specificity, 67% sensitivity, and an AUC/ROC value of 0.89, for a 0.4 decision threshold. Our ML algorithm effectively detects abnormal LVG even at early stages. Innovative solutions are needed to improve risk stratification of patients without established CVD, and ML may enable progress in this direction.

The heart beats are due to the synchronized contraction of cardiomyocytes triggered by a periodic sequence of electrical signals called action potentials, which originate in the sinoatrial node and spread through the heart’s electrical system. A large body of work is devoted to modeling the propagation of the action potential and to reproducing reliably its shape and duration. Connection of computational modeling of cells to macroscopic phenomenological curves such as the electrocardiogram has been also intense, due to its clinical importance in analyzing cardiovascular diseases. In this work, we simulate the dynamics of action potential propagation using the three-variable Fenton–Karma model that can account for both normal and damaged cells through a the spatially inhomogeneous voltage diffusion coefficient. We monitor the action potential propagation in the cardiac tissue and calculate the pseudo-electrocardiogram that reproduces the R and T waves. The R-wave amplitude varies according to a double exponential law as a function of the (spatially homogeneous, for an isotropic tissue) diffusion coefficient. The addition of spatial inhomogeneity in the diffusion coefficient by means of a defected region representing damaged cardiac cells may result in T-wave inversion in the calculated pseudo-electrocardiogram. The transition from positive to negative polarity of the T-wave is analyzed as a function of the length and the depth of the defected region.

Left ventricular hypertrophy (LVH) is classified as concentric or eccentric based on left ventricle relative wall thickness. Using machine learning (ML) techniques on basic clinical parameters and features from a single-lead electrocardiogram (ECG) we detected LVH in a hypertensive population without cardiovascular disease. We enrolled 812 hypertensive subjects with no indications of cardiovascular disease. Based on left ventricular mass index (LVMi) and relative wall thickness (RWT), the subjects were classified into 2 groups; i) normal geometry (NG) and concentric remodeling (CR) were classified as no-LVH, whereas ii) concentric and eccentric hypertrophy were classified as having LVH. We trained a Random Forest to distinguish between the two categories. For comparison we also trained a logistic regression and a convolutional neural network model. We performed feature importance and interaction analysis using SHAP to interpret the model's predictions for feature importance and feature interactions. Our model was able to distinguish no-LVH subjects from the ones with LVH with ROC/AUC 0.82 (95% CI: 0.71-0.91) and average precision 0.61. At threshold 0.3, specificity is 81% and sensitivity is 58%. Age, corrected QT interval, T wave duration, and being female, were the most important features that contributed to the model's predictions. Utilizing analysis of single-lead ECG combined with clinical data, we demonstrated strong performance in detecting LVH. The identification of key ECG features such as corrected QT interval and T wave morphology underscores the clinical value of single-lead ECG analysis. These findings are particularly important in the era of wearable devices, where accessible, noninvasive screening for cardiac conditions like LVH can be integrated into everyday health monitoring.

Cardiac remodeling is recognized as an important aspect of cardiovascular disease (CVD) progression. Machine learning (ML) techniques were applied on basic clinical parameters and electrocardiographic features for detecting abnormal left ventricular geometry (LVG), even before the onset of left ventricular hypertrophy (LVH), in a population without established CVD. After careful screening, we enrolled 528 subjects with and without essential hypertension, but no other indications of CVD. All patients underwent a full echocardiographic evaluation and were classified into 3 groups; normal geometry (NG), concentric remodeling without LVH (CR), and LVH. Abnormal LVG was identified as increased relative wall thickness (RWT) and/or left ventricular mass index (LVMi). We trained nonlinear predictive ML models, to classify subjects with abnormal LVG and calculated SHAP values to perform feature importance and interaction analysis. Hypertension, age, body mass index over the Sokolow-Lyon voltage, QRS-T angle, and QTc duration were some of the most important features. Our model was able to distinguish NG from all others (CR+LVH), with accuracy 86%, specificity 75%, sensitivity 95%, and area under the receiver operating curve (AUC/ROC) 0.89. We also trained our model to classify NG and CR (NG+CR) against those with established LVH, with accuracy 89%, specificity 97%, sensitivity 50%, and AUC/ROC 0.85. Our ML algorithm effectively detects abnormal LVG even at early stages. Innovative solutions are needed to improve risk stratification of patients without established CVD, especially in primary care settings, and ML may enable this direction.

Background: Hypertension is a major risk factor for cardiovascular disease (CVD) which often escapes the diagnosis or should be confirmed by several office visits. The electrocardiogram (ECG) is one of the most widely used diagnostic tools and could be of paramount importance in patients’ initial evaluation. Methods: We used machine learning (ML) techniques based features derived from the electrocardiogram for detecting hypertension in a population without CVD. We enrolled 1091 subjects who were classified into hypertensive and normotensive group. We trained a random forest (RF), to predict the existence of hypertension in patients based only on a few basic clinical parameters and ECG-derived features. We also calculated Shapley additive explanations (SHAP), a sophisticated feature importance analysis, to interpret each feature’s role in the random forest’s predictions. Results: Our RF model was able to distinguish hypertensive from normotensive patients with accuracy 84.2 %, specificity 66.7%, sensitivity 91.4%, and area under the receiver-operating curve 0.86. Age, body mass index (BMI), BMI-adjusted Cornell criteria (BMI multiplied by RaVL+SV ₃ ), R wave amplitude in aVL, and BMI-modified Sokolow-Lyon voltage (BMI divided by SV ₁ +RV ₅ ), were the most important anthropometric and ECG-derived features in terms of the success of our model. Conclusions: Our ML algorithm is effective in the detection of hypertension in patients using ECG-derived and basic anthropometric criteria. Our findings open new horizon in the detection of many undiagnosed hypertensive individuals who have an increased cardiovascular disease risk.