A Framework for Ischemic Beat Detection using Multi-Layer Feedforward Neural Network and Principal Component Analysis (IBD-MLFFNN-PCA)

In this paper, we describe a novel procedure to automatically classify the ischemic beats of an ECG signal using multilayer feedforward neural network (MLFFNN) and principal component analysis. A framework for the proposed method is also suggested. ECG recordings are given as input to the system and ischemic beats are detected from the recordings. The efficacy of the framework is tested using recordings from the European ST database. The ECG recording goes through preprocessing steps and feature vector formation. The feature vector is further reduced using principal component analysis. This reduced feature vector is given as input to the 3 layer MLFFNN which uses scaled conjugate gradient backpropagation algorithm for training. The network was further validated and tested to check its efficacy. The efficacy of the procedure was measured using sensitivity (91%) and positive predictive accuracy (94%). Also the proposed method is compared with some other existing work and it is seen that the proposed method outperforms others.