Extraction of fetal ECG using adaptive Volterra filters

In this paper we present a new method for extracting the fetal electrocardiogram (FECG) signal from one thoracic ECG signal and one or more abdominal signals. Our method is based on the use of an adaptive Volterra filter (AVF) that is capable of synthesizing the nonlinear relation between the mother thoracic ECG signal and the abdominal signals which contains a transformed mother ECG, the fetal ECG and other noise elements. An adaptive multi-sensory noise canceler structure is adopted for the extraction purpose. In the case where more than one abdominal signals are used, the proposed algorithm uses a linear combiner (LC) to form a primary signal from those abdominal signals. The LC and the AVF are updated by the RLS algorithm. The proposed method is applied to real ECG measurements to demonstrate its superior effectiveness.