Heartbeat Classification Based on PCA and CNN

This work presents a method for automatic heartbeat classification based on principal component analysis and a convolutional neural network on ECG signals. We developed a database holding the first ten principal components and the relative RR intervals of P-QRS complexes from the MIT-BIH Arrhythmia Database patients. The convolutional neural network was used to obtain a model for classifying heartbeats based on this database. This model was tested and compared to other algorithms existing in the literature, and the results evidenced the relative advantages of the method.