An Efficient Soft-Computing Technique for Extracting Fetal ECG from Maternal ECG Signal

The Fetal Electrocardiogram (FECG) signal reflects the electrical activity of the fetal heart. It contains information about the health status of the fetus and as a result, an early diagnosis of any cardiac defects before delivery increases the effectiveness of the appropriate treatment. The proposed approach extracts the FECG from two ECG signals recorded at the thoracic and abdominal areas of the mother, with the help of a hybrid soft computing technique. The thoracic ECG is assumed to be almost completely maternal (MECG) while the abdominal ECG is considered to be composite because it contains both the maternal and the fetal ECG signals. The principle used for the elimination of artifacts is ANC. An Adaptive Neuro-Fuzzy Inference System (ANFIS) is used to remove the artifacts and to extract the FECG component from abdominal signals of very low maternal to fetal signal-to-noise ratio. After removing the artifacts using ANFIS, better results are obtained by optimizing the ANFIS parameters using Swarm Intelligent Technique, namely Particle Swarm Optimization (PSO). The experimental results show that the proposed approach can effectively remove artifacts and extract the desired FECG signals from the abdominal signals.