ECG Signal Classification Using Continuous Wavelet Transform and Deep Neural Network

Accurate classification of cardiac activity using electrocardiogram (ECG) signals is crucial in preventing and treating many heart-related diseases. However, the complex and non-stationary nature of ECG signals poses challenges for traditional classification methods. This paper presents an effective approach that combines Continuous Wavelet Transform (CWT) for feature extraction with a Deep Neural Network (DNN) for classification. The CWT is used to transform one-dimensional ECG signals into two-dimensional time-frequency spectrograms, preserving both spectral and temporal characteristics. The proposed method utilizes CWT to transform raw ECG signals into time-frequency representation maps, preserving both spectral and temporal features. This transformation reduces the need for complex pre-processing while enhancing classification accuracy. The proposed method was evaluated on benchmark datasets, achieving superior performance compared to traditional techniques in terms of accuracy, precision, recall, and F1-score. The integration of CWT and DNN demonstrates its potential to enhance diagnostic accuracy in real-world applications, making it a promising tool for clinical use.