Construction and validation of a diagnostic model for interventricular septal hypertrophy based on electrocardiographic parameters and clinical factors

Interventricular septal hypertrophy (IVSH) is considered to be an early-stage alteration of left ventricular hypertrophy and is a major risk factor for a wide range of fatal arrhythmias and heart failure. We aimed to develop and validate a diagnostic model to detect patients with IVSH. This study included 1097 participants grouped according to the presence or absence of IVSH. An echocardiography (UCG) was used as a diagnostic criterion for IVSH, and we fit three logistics regression models, each composed of (a) electrocardiographic parameters (ECG-IVSH), (b) clinical factors (Clinic-IVSH), and (c) terms for both ECG-IVSH and Clinic-IVSH (Joint-IVSH). We assessed the performance of the models by calculating discrimination (area under the receiver operating characteristic curve, AUROC), calibration (calibration curves), and clinical value (clinical decision curves). The SHapley Additive exPlanations (SHAP) method explained global feature importance. The Joint-IVSH, which consisted of V 1 lead R-wave duration (V 1 Rd), V 1 lead R-wave area (V 1 Rs), age, body surface area, hypertension history, smoking status, and drinking status, had the highest accuracy compared to the other two models, with sensitivity and specificity of 76.7% and 82.1% after analysis of ROC curves (AUC 0.859, 95% CI 0.826, 0.880), consistency between the diagnostic results and the actual results, and the favorable clinical validity of the model in diagnosing IVSH. SHAP analysis prioritized V 1 Rd as the top contributor. Joint-IVSH performs well in discriminating IVSH events and can provide valuable reference information to guide clinical judgment and therapeutic schemes.