Interpretable hybrid model for an automated patient-wise categorization of hypertensive and normotensive electrocardiogram signals

•Interpretable hybrid model for patient-wise electrocardiogram (ECG) categorization as normotensive and hypertensive is proposed.•Proposed architecture consists of convolutional neural network (CNN) and support vector machine (SVM) classifier as hybrid model.•In addition, an automated digitization algorithm is proposed for converting paper-ECG to 1D digital signal.•Achieved highest accuracy of 93.33%, sensitivity of 100% and specificity of 87.5% in patient-wise classification.