Many studies focus on designing automatic Electrocardiogram (ECG) disorder detection algorithms with the growing development and deployment of deep learning-based systems. Previous literature evidence that many deep learning algorithms achieve high accuracy in automatic ECG disorder detection. However, three main challenges remain evident: (1) clinical ECG datasets are expensive, (2) there need to be more expert annotations for the available datasets, and (3) there need to be more rare CVDs covered by datasets. Therefore, we introduce a novel methodology for generating synthetic ECG signals using a mathematical model and a bio-inspired algorithm to estimate the model’s parameters. We expand an existing mathematical model to reproduce different cardiac arrhythmias using ECG records as a reference. The model’s parameters differ for each selected ECG wave, obtained by minimizing the difference between the ECG and synthetic recordings. Our results show that the proposed methodology can estimate the synthetic ECG model parameters using Gaussian functions for each ECG wave, which is highly adaptable to different cardiac diseases.