A tutorial review on data compression with detection of fetal heart beat from noisy ECG

FECG (Fetal Electrocardiogram) signal convey information in making timely decisions during pregnancy. It is very important to extract and detect the FECG signal from a composite maternal abdominal signals for task of fetal monitoring. We implement adaptive filters for the task of separation of FECG signal from the abdominal ECG signal and understand its nature. A Least Mean Square (LMS) algorithm is implemented for the purpose of detection adaptively. Data compression techniques are implemented foe the filtered ECG signal for the purpose of easy storage and transmission.

Since the data is more important and sensitive, it is necessary to implement lossless compression with high compression ratio by reducing the redundancy involved in the ECG signal with low noise. A Fast Fourier Transform, Discrete Cosine Transform (DCT), DCT-II and Discrete Sine Transform (DST) methods are used here for the purpose of compression. The performance metrics like compression ratio, low percent mean square (PRD) are compared and the results are tabulated by using Matlab software tool.

Share This Post