Abstract: Routine fetal heart beat (FHB) clinical detection method is the fetal electrocardiogram (FECG) signal detection， which needs to attach multi-electrodes on the maternal body， limits the recordings during labor and causes inconvenience. Therefore， a non-invasive FHB detection method using fast independent component analysis (FastICA) algorithm was proposed. Through analyzing the principal component of maternal ballistocardiogram(MBCG)， FHB detection can be implemented without attaching electrodes to the subject， which is suitable for home unobtrusive health care. 45 groups of MBCG signals and maternal electrocardiogram (MECG) signals were collected synchronously to conduct the accuracy 91.3% .

Key words: fetal heart beat signal, maternal ballistocardiogram signal, fast independent component analysis algorithm, fetal electrocardiogram signal, maternal electrocardiogram signal