ECG Beat Recognition using Principal Components Analysis and Artificial Neural Network

The analysis of heart beat cycles in an ECG (electrocardiogram) signal is essential for long-term monitoring of heart patients. However, it is a very tedious and time-consuming task to analyze the ECG recording beat by beat in a long-term monitoring. This is because the abnormal heart beats can occur randomly and a long-term ECG record, say 24 hours, may contain hundreds of thousands of beats. Hence, it is highly desirable to automate the entire process of heart beat classification. The present work proposes a technique for heart beat recognition. First, the individual beats belonging to each category were extracted from the MIT-BIH arrhythmia database using an R-peak detection algorithm and after preprocessing, features are extracted from the beats using Principal Component Analysis (PCA) .This process drastically reduces the dimensionality of the vectors to be classified. The feature vectors thus obtained are used to train a neural network (NN) classifier. After the network is trained, its performance in terms of its generalizing ability is tested on a separate test dataset which was not used during training.