Investigations on performance analysis Of noise classification techniques with Fir filtering in fetal electrocardiogram And transceiver using vlsi Implementation

Abstract

Electrocardiogram (ECG) is the non-invasive diagnostic tool for providing potential data about cardiac disorders of both maternal and fetus through computer-aided design by placement of surface electrodes. As per the data collected from Indian Paediatrics 2018, it is reported that 200,000 babies are born every year with Coronary Heart Disease (CHD) out of which one-fifth are having severe defects due to negligence and non-identification of risk situations during the neonatal phase of the baby birth. Hence, it is crucial to diagnose the Fetal ECG (FECG) signal using monitoring devices but unfortunately during which the noise intrusion imbibes major cause to the observations recorded. Therefore, the overall scope of the thesis is to devise a filter upon extracting the required features of FECG on applying digital signal processing techniques for further clinical processing involving classifiers and also to enhance the performance of transceivers. newlineFetal Heart Rate (FHR) monitoring has challenges during diagnosing the health status due to noise sources either internal or external, thereby causing the forthcoming interpretation of results to deteriorate. Hence, the major contributions in this work are to classify and identify the five different types of noise (Powerline Interference, Muscle Contraction, Electrode Contact, Patient Movement and Electrosurgical) at high accuracy and noise sensitivity with the proper choice of QRS peak detection algorithms and classifiers. The data utilized for the purpose is taken from MIT- BIH Arrhythmia database. On interpreting the noise type, an appropriate programmable FIR filtering technique with reduced delay and power are designed using VLSI tools for different windowing methods newline