The document presents a method for classifying ECG signals using continuous wavelet transform (CWT) and deep neural networks. CWT is used to decompose ECG signals into different time-frequency components, which are then used to generate a scalogram image. A convolutional neural network is used to extract features from the scalogram images and classify the ECG signals into types including ARR, CHF, and NSR. The method achieves classification accuracy of over 98% on a public ECG dataset, outperforming other methods. The simple and accurate approach has potential for use as a clinical diagnostic tool.