QRS complex characterization based on non-negative matrix factorization NMF

In this work a novel automated method is presented for the detection and characterization of the QRS complexes of the fetus cardiac activity by using single channel maternal ECG recordings. The proposed algorithm is based on the continuous wavelet transform CWT and non-negative matrix factorization NMF. Clinically significant information in the fetal electrocardiogram signal is often masked by maternal electrocardiogram mECG considered as the most predominant interference. The aim of this work is to characterize the complex without being obliged to extract the fetal signal from that of the mother. We focus on both time domain and frequency domain to extract the useful characteristics of the signal.