J. P. Saini

Design and Comparison of Digital Filters for Removal of Baseline Wandering from ECG Signal

Heart diseases, which are one of the death reasons, are among the several serious problems in this century and as per the latest survey, 60% of the patients die due to Heart problems. These diseases can be diagnosed by ECG (Electrocardiogram) signals. ECG measures electrical potentials on the body surface via contact electrodes thus it is very important signal in cardiology. Different artifacts affect the ECG signals which can thus cause problems in analyzing the ECG Thus signal processing schemes are applied to remove those interferences. The work proposed in this paper is removal of low frequency interference i.e. baseline wandering in ECG signal and digital filters are designed to remove it. The digital filters designed are FIR with different windowing methods as of Rectangular, Gaussian, Hamming, and Kaiser. The results obtained are at a low order of 56. The signals are taken from the MIT-BIH database which includes the normal and abnormal waveforms. The work has been done in MAT LAB environment where filters are designed in FDA Tool. The parameters are selected such that the noise is removed permanently. Also the best results are obtained at an order of 56 which makes hardware implementation easier. The result obtained for all FIR filters with different windows are compared by comparing the waveforms and power spectrums of the original and filtered ECG signals. The filters which gives the best results is the one using Kaiser Window.

Design of Digital IIR Filter for Noise Reduction in ECG Signal

Electrocardiogram (ECG) signal has been widely used in cardiac pathology to detect heart disease. A digital infinite-impulse response (IIR) filter design is proposed in this paper. This includes an implementation and evaluation of butter worth low pass infinite impulse response filter method to remove high frequency noise and for this filter is applied to noisy ECG data sample and original sample are taken as reference signal. The suggested method considers the magnitude response for choosing the cutoff frequency and the FFT spectrum estimate response to find the lowest filter order. The structure and the coefficients of the digital IIR filter are designed using FDA tool in MATLAB. The filter outputs average power before and after filtration are calculated using FFT and for simulation of this filter, the hardware is designed using micro controller At mega 16 A. For hardware designing the samples taken are record no. 108 and record no. 119 (taken from MIT-BIH database, ML II signal). Here samples are taken from MIT-BIH arrhythmia database (mitdb) ML II are used.

Abnormality Detection in Indian ECG using Correlation Techniques

The paper proposes a method based on signal processing correlation technique to find out whether the ECG is normal or abnormal. Many of the abnormal ECGs are called Arrhythmias. ECG (lead II) obtained from conventional ECG machine of Indian patients are digitized and the data are crosscorrelated with the reference standard normal ECG data. Two different beats of the same ECG data are also correlated. The correlation parameters are used to identify the ECG as normal or abnormal. The accuracy obtained in this method is 100%. The cross-correlation is done using MATLAB 7.12.0 (R2011a) tools.

Extracting Samples as Text from ECG Strips for ECG Analysis Purpose

The paper presents a work of conversion of ECG signal from ECG strips/papers to samples as text which is analogous to the standard MIT-BIH Normal Sinus Rhythm/Arrhythmia database. ECG strips are scanned and then using MATLAB, the data is obtained for the ECG taken from Indian patients. The parameters like Heart rate, PR interval, QRS duration, QT interval and RR interval are compared. The result shows 99% accuracy in the data obtained by this method. The process of extraction of ECG data is also validated for Normal Sinus Rhythm/Arrhythmia from MIT-BIH database. The correlation obtained is nearly 98%. Thus the method is useful for automatic analysis of ECGs from ECG strips at rural areas also.

Comparison of Different Digital Filters for QRS Complex Extraction from Electrocardiogram

In this paper a method for real time QRS complex extraction from Electrocardiogram using digital filter is proposed. The digital filters are used to detect the different signal features on an human heart electrocardiogram signal. The waveform feature of interest in this paper is the QRS complex of electrocardiogram signals taken from MIT-BIH Arrhythmia Database. The digital filters designed are FIR which are of a order 20 which is quite less. A comparison of different FIR digital filters based on Rectangular, Gaussian and Kaiser windows is done. All the work has been done with MATLAB®. The performance of digital filter is described by the comparison of mean square error (MSE) of QRS complex extracted, for three different patient records of MIT-BIH, by our method using different windows and QRS complex of referred original ECG recording of same MIT-BIH records. The MSE for Kaiser Window was minimum.

Uniform Sampling of ECG Waveform of MIT-BIH Normal Sinus Rhythm Database at Desired Intervals

MIT-BIH Database is the standard ECG database which is used universally for ECG analysis purpose. MIT-BIH database for normal sinus rhythm is sampled at 128 Hz and the data is available at uniform intervals of 7.8125 ms. To use this data for analysis purpose with various techniques like artificial neural networks, correlation techniques etc., it is required to have samples at desired intervals. Hence this paper proposes an image processing method to convert the samples at desired intervals, so that the MIT-BIH database can be used widely and universally. General Terms ECG uniform sampling

Fault Diagnosis of RC-coupled Amplifier Using Slope Fault Feature and Comparision with Different Neural Networks

This paper describe fault diagnosis of RC-Coupled amplifier using slope fault feature. These slope fault feature technique utilized to construct the fault dictionary for RC-Coupled amplifier. This fault dictionary used to generate different fault diagnosis model for analog circuit using artificial neural network technique. For generate the fault model three different type neural networks utilized. These neural networks are radial basis function neural network, perceptron neural network and feed forward back propagation algorithm neural network. In theses network radial basis function neural network shows 100 percentage efficiency, perceptron neural network shows 87.5 percentage efficiency and feed forward back propagation algorithm shows 99.31 percentage efficiency in the training and testing for fault dictionary.

Simulation of Individual Electrical Potentials of Heart at Arms and Legs

Sinoatrial (SA) node is the natural pacemaker which triggers itself and the signal is passed over to other nodes of the Heart. Electrocardiogram (ECG) is the result of these electrical activities which is generally picked up by the electrodes from the arms and legs. This paper explains the procedure to find the potentials at arms and legs from the augmented leads graphs taken from any standard ECG machine. MATLAB tool is used to extract the data from the augmented lead graphs. First the graph is scanned and then using image processing tools of MATLAB, the data is extracted from augmented leads. It is scaled by a factor of 1.5 and the signals at arms i.e. RA, LA and leg i.e. LL are generated. Lead II signal obtained from this simulated arms and legs signals is compared with the Lead II graph obtained from the ECG machine of the same patient. The result of correlation obtained with the simulated Lead II signal and Lead II signal taken directly from the ECG machine is 100%, which shows that using the above technique, the electrical activities of the SA node are simulated properly and the individual potentials of the Heart obtained at arms and legs can be used for further analysis purposes.