Processing and analysis of foetal phonocardiograms

A technique for segmenting the foetal phonocardiogram into the systolic and diastolic phases on a per beat basis is presented. All the parameters of cardiac function are thereby derivable. Contrary to other methods of automated segmentation, the system uses only one transducer and no other input. The technique makes no assumptions about the foetal heart rate or the beat-to-beat variability. The system relies on external knowledge and contextual information to aid signal analysis. System performance is resistant to high levels of noise and missing heart sounds. Accuracy is good