Title :
MUSIC-based bispectrum detector: a novel non-invasive detection method for overlapping fetal and mother ECG signals
Author :
Zgallai, Walid ; Sabry-Rizk, Madiha ; Hardiman, Paul ; O´Riordan, John
Author_Institution :
Sch. of Eng., City Univ., London, UK
fDate :
30 Oct-2 Nov 1997
Abstract :
This paper describes a novel detection method which utilises some 3-d unique bispectral features peculiar to mother and fetal transabdominally measured ECG signals to detect the number of occurrences of fetal heart beats within the maternal cycle frame before and during labour. The bispectrum detector requires the MUSIC super resolution spectral estimation to pin point different resonant frequencies associated with the QRS waves in each of the mother and fetal signals. Results obtained from 15 preoperative labour cases using both non-invasive “transabdominal” and fetal scalp electrodes show that the MUSIC detection in the second-order spectral domain can recover most of the fetal heart beats particularly those occurring at the same time as the mother´s. The MUSIC also enhances fetal and mother ECG resonances and their harmonics. However, at a given false alarm probability due to poor signal-to-noise ratios the bispectrum of the MUSIC estimator has shown a remarkable performance not only in discriminating against impulsive non-Gaussian and transient noise but also against the mother ECG relatively strong harmonics at the fetal QRS frequencies
Keywords :
electrocardiography; medical signal detection; medical signal processing; obstetrics; spectral analysis; MUSIC-based bispectrum detector; QRS waves; electrodiagnostics; false alarm probability; fetal heart beats; fetal scalp electrodes; impulsive nonGaussian noise; labour; maternal cycle frame; noninvasive detection method; overlapping fetal/maternal ECG signals; polyspectrum; transabdominally measured ECG signals; transient noise; Detectors; Electrocardiography; Fetal heart; Frequency estimation; Multiple signal classification; Resonant frequency; Scalp; Signal detection; Signal resolution; Signal to noise ratio;
Conference_Titel :
Engineering in Medicine and Biology Society, 1997. Proceedings of the 19th Annual International Conference of the IEEE
Conference_Location :
Chicago, IL
Print_ISBN :
0-7803-4262-3
DOI :
10.1109/IEMBS.1997.754466