An automated methodology for FECG extraction and Fetal Heart Rate monitoring using Independent Component Analysis

This paper introduce an automated methodology for the extraction of Fetal Electrocardiogram (FECG) from Abdominal electrocardiogram (AECG) recording. FECG is a weak signal from the maternal ECG indirectly measured by surface electrode placed on mother´s abdomen. The Fetal signals are buried in other interference signal. Extracting FECG from the strong background interference has an important value in clinical application. Since, lots of research work has been perform in this field, some of these are Threshold and Filtering method, Neural Network method, Wavelet Transform and other. The proposed approach involves Independent Analysis technique for FECG extraction. Basically ICA works in different parameter Kurtosis and Negentropy. Here we are focusing on Negentropy, after extraction of FECG signal Fetal R-peak are located using Threshold free detection method which involves R-R moving interval, calculated on the basis of normal maximum and minimum heart rate. This algorithm is implemented on 20 recorded signals using MATLAB. The average accuracy and average positive Predictivity of the detection method are 97.62% and 98.98% respectively.