Investigation on Maternal and Fetal Heart Signal Extraction Using Adaptive Filtering Techniques and Integrating with Block Sparse Bayesian Learning

This paper consists of the implementation of fetal electro cardiogram extraction from mother's womb. Electro cardiogram test is very important to know about fetal condition so as to avoid stillbirth and neonatal death during pregnancy. The fetal ECG can be extracted from a woman's abdomen and thorax which consist of multiple unwanted noises and artifacts. The techniques used for extracting the FECG are ADAPTIVE NOISE CANCELLATION and COMPRESSION SENSING. Adaptive noise cancellation uses least mean squares (LMS), normalized least mean squares (NLMS), leaky least mean square (L-LMS) algorithms. Block sparse Bayesian learning via bounded optimization (BSBL-BO) algorithm is used to reconstruct non-sparse signals.