Abstract: In this paper an approach is presented to analyze the fetal health through the study of fetus Electrocardiogram (ECG). This paper presents the detail survey on different features of fetal heart which are P-wave, QRS wave and T-wave amplitude and width, segment intervals like PQ interval, RR interval, Heart rate, Heart rate variability, fetal heart axis, these are the standard and most required features. This paper reviews the different features of fetal heart and different feature extraction techniques through fetal electrocardiogram and thus classifying the fetus into normal and abnormal class.

In this paper a hybrid approach is presented to study the fetus ECG. Processing the ECG includes pre-processing, feature extraction, beat detection and classification. Here, we have implemented Pan tompkin algorithm for feature extraction, and PTE algorithm for beat detection. According to the literature, classification accuracy claimed is in the range of 92 to 99.68%. Though there has been significant variation in the range of accuracy it is at the cost of various other factors such as computational complexity. For simulation, Abdominal and direct Fetal ECG dataset of various time samples of different frequencies are taken from physionet.org database.

Keywords: Fetal ECG, preprocessing, feature extraction, classification, fetus features. 

| DOI: 10.17148/IJIREEICE.2020.8632