International Journal of Industrial Electronics and Electrical Engineering, ISSN: 2347-6982 Volume-4, Issue-9, Sep.

-2016

PARAMETERS CALCULATION AND CLASSIFICATION OF ECG

SIGNALS USING LAB VIEW
1
DEEPA PAUL, 2PRIYANKA CHOUDHARY, 3RAM AVTAR JASWAL
1,2
P.G Scholar, Electrical Engineering, UIET Kurukshetra University, Hrayana, INDIA
3
Assistant Professor, Electrical Engineering, UIET Kurukshetra University Hrayana, INDIA
E-mail: [email protected], [email protected]

Abstract— “Why are we using LABVIEW to study ECG in a better fashion?” Keeping this question in mind, we wrote this
paper contract with study and performance of ECG signal processing in LABVIEW with biomedical toolkit 2013. The raw
ECG is attain from physio.net MIT-BIH Data base. These tools can also help us to denoising, detrending, and extracting
ECG signal. Study of ECG signals of different filtering techniques includes WA De-trend Express VI, Wavelet De-noise,
Wavelet packet from wavelet families. Besides, Features are extracted with LABVIEW as a successful way to calculate the
parameters of ECG signal more expediently. LABVIEW with its signal processing prove that it is easy and strong for
resolving ECG signal processing problems as compared to MATLAB software.

Keywords— ECG parameters, LABVIEW Biomedical Toolkit, MIT-BIH data base, WA De-noise , Wavelet transform.

I. INTRODUCTION II. METHODOLOGY

ECG (Electro-Electrical Signals, Cardio-Heart,
Gram- Graph) is an electrical activity or movement of LABVIEW is more effective and easy to use because
heart which is recorded by placing electrode on signal processing related toolkits can provide an
human body at different parts which is shown in efficient environment and tools for ECG signal
graph is called electrocardiogram (ECG). ECG signal analysis.
is non-stationary signals, hence it changes with time
so it very difficult to proper diagnosis the disease of
heart patients. An ECG signals represents heartbeats
consisting of P Wave, QRS Wave, T Wave and U
Wave. The amplitude, duration and interval between
each of these waves help us distinguish between
normal and abnormal waveforms. The ECG ranges
from a few µV to about 1V in magnitude. Whereas
the characteristic waves of an ECG have a maximal
magnitude of only few mV, a wandering baseline in
the ECG due to variations in electrode-skin
impedance may reach 1V. The amplifier Bandwidth
is commonly between 0.05 and 100-500Hz [1]. The
amplitude of a wave is measured with reference to the
ECG baseline level, commonly defined by the Iso-
electric line which immediately come first the QRS
complex. The human heart consist of four chambers
two Atria called upper chambers and two ventricles Figure 2. Steps of ECG signal analysis
called lower chambers. Normally the heart begins at
the right atrium called sinus artia (SA) node and a In these paper we work on five different phase which
special group of cells transmits electrical signal has been shown by flowchart of general steps of ECG
across the heart. The standard ECG waveform for one signals in figure 2.
cardiac cycle in figure 1 as shown below : In above flow chart firstly, ECG signal is acquire then
pre-processing has been done by removing base line
wandering and wide band noise , After that
extracting the features from the denoise the ECG
signal and at last calculate the each parameters of
ECG signal.

ECG SIGNAL ACQUISITION

Generally, ECG signal is acquired from MIT-BIH
data base or from www.physionet.org.
Physionet>>physiobank>>physiobankATM>> MIT-
BIH Data base. The acquisition of ECG signal using
Figure 1 one cardiac cycle of ECG signal LABVIEW [2] shown in figure 3.
Parameters Calculation and Classification of ECG Signals Using Lab View

63
International Journal of Industrial Electronics and Electrical Engineering, ISSN: 2347-6982 Volume-4, Issue-9, Sep.-2016

Following techniques are used to remove the noisy

signal or wide band noise but we were use only two
technique to get the signal noise free as discussed
below:
A) Wavelet De-noise.
B) Wavelet Packets.

A) Wavelet De-noise: ECG signal are stained with

Figure3. Acquiring ECG signal
noise which make unproper diagnosis to determine
ECG signal Pre Processing any type of diseases. This types of noise could not be
Generally, the recorded ECG signal is frequently removed by traditional filters so Wavelet De-noise
contaminated by the noise and other artifacts which Express VI is used to decompose the ECG signal.
destroy the useful information (normally frequency Firstly, These signals are divided into several sub-
components) from signals. Pre-Processing of ECG band which depends upon the wavelet level as u give
signals helps us to remove this contaminated from the to your signal then it modify each coefficients of
ECG signal such as : wavelet by applying soft threshold function, single
1. Base line wandering . level with db06 wavelet. Undecimated Wavelet
2. Wide band noise. transform (UWT) gives better results as compared to
3. Power line interference FFT or DWT which is also used to remove the
4. Electromyographic (EMG) noise wideband signal [5,6].
5. Electrode pop noise The typical shape of recording electrical activity of
6. Patient-electrode motion noise heart is well known to cardiologistic. Any significant
Among of all these noises the baseline wandering and derivation from that shape is usually considered to be
power line interference can strongly affect the ECG a symtoms of a pathological condition. A
signal analysis. The following methods can be used to pathological condition can sometime be diagnosed
remove these noises . more easily when frequency content of the signal
1. Removing Baseline Wandering: It usually comes analysed. So Wavelet de-noise is used for this
from respiration at frequencies wandering between purpose.
0.15 to 0.3 Hz. There are two different methods to
remove this type of methods: Digital high pass filter
and Wavelet transform.
In our work Wavelet transform are used to reduced
the low frequency trend of signal as WA de-trend VI
are using in it, after removing baseline wandering the
ECG signal get more stationary signal [3].

Figure 5 Denoising the ECG signal using Wavelet De-noise

Figure 6 shows setting of wavelet denoise in which

transform type = UWT; levels = 8; wavelet db06 ;
soft thresholding is used to show every single points
and threshoding level= single .

Figure 4 Base line wandering removal using WA de-trend

express VI
Therefore, Trend signal is estimated by using
following formulae:
Trend level = log2(2*T)/ log2(N)
Where T is the sampling duration and N is the no of
sampling points in raw ECG signal, we use 60
sampling duration and 12000 sampling points in total
so that the trend level of threshold frequency is of 0.5
[4].

2. Wide band noise: After removing the base line

wandering the resulting ECG signal is become
Figure 6. Wavelet transform EXPRESS VI for denoise the
stationary as compared to the raw ECG signal. ECG signals

Parameters Calculation and Classification of ECG Signals Using Lab View

64
International Journal of Industrial Electronics and Electrical Engineering, ISSN: 2347-6982 Volume-4, Issue-9, Sep.-2016

B) Wavelet Packets: It specifies the wavelet type

with discrete wavelet transform analysis and
Undecimated wavelet transform for denoising the
signal. The wavelet packets are generalization of
wavelet transform that allow for arbitrary tree-shaped
band pass filtering. A set of detail and approximation
components of signal is called wavelet packets Figure 9 Parameters of ECG signals
decomposition tree. In order to apply wavelet packet
analysis let us define the scaling function ( )= Above figure calculate each Parameters of ECG
( ) signal includes as Heart Rate mean , Heart rate std ,
and the wavelet function ( ) = ( ). QRS amplitude mean ,QRS amplitude std, QRS Time
mean , QRS Time std, PR interval mean , PR interval
Then it can be written functions ( ), = std, QT interval mean , QT interval std, ST level
0,1,2 … ., as mean , ST level std, Iso level mean, Iso level std. as
( ) = 2 ∑ ℎ( ) (2 − ) i shown in table 1
( )= 2 ∑ ( ) (2 − ) ii Often times, feature extraction of ECG gives a proper
Equation i and ii are analysis functions called diagnosis of cardiac problem and its help to identify
wavelet packet atoms [7,8]. the condition of patient . The results of decomposed
signal will show the important details and
approximation coefficients, which represents the
original signal.

Table 1 ECG parameters and their normal values

Figure 7 Denoise of ECG signal using wavelet packet

In figure below shows the setting regarding technique

wavelet type is wavelet packet ; level is 4; Entropy
type is threshold ; optimal parameter = 0.5 .

CLASSIFICATION OF ECG SIGNAL

After calculate the ECG parameters we will classify
the disease based on interval and feature extraction to
describe the ECG signal with P-QRS-T wave and its
interval. There are different types of heart disease are
shown in table 2.

Table 2 Classification of heart disease

Figure 8 Denoise of the ECG signals using wavelet packet.

Features Extraction and ECG Parameter

Calculation
Features information of ECG signal helps in
diagnosis of cardiac diseases. We frequently need to
extract many features from the pre-processed or III. EXPERIMENTAL RESULTS
denoised ECG signal for the diagnosis purposes it
includes P onset, P offset, QRS onset, QRS offset, T Simulation is done by using LABVIEW biomedical
onset, T offset after that we calculate the ECG toolkit where each and every components are
parameters such as heart rate, PR interval, ST level, describe and explained in proper way. Each parts of
QRS amplitude, QRS intervals, ISO level etc. simulation model are available easily in LABVIEW
Parameters Calculation and Classification of ECG Signals Using Lab View

65
International Journal of Industrial Electronics and Electrical Engineering, ISSN: 2347-6982 Volume-4, Issue-9, Sep.-2016

VI so that we can add, remove and replace with After De-trend the noisy signal, wavelet de-noise is
another one. After acquiring the raw ECG signal from done by applying directly from tool kit of labview in
MIT-BIH database [13] then the raw signal which is which signal processing tool shows different types of
contaminated by noise is going to de-trend and trend wavelet you can apply any one wavelet transform
level for boost the signal from zero value ( baseline ) from wavelet families. Denoising of continued signal
and that variation between baseline and staring of raw is now clear the noise which is easily shown on figure
signal is called Baseline wandering. This type of is called power line interference. Generally, PI
noise is removed by Wavelet De-trend VI. interference is arise power line disturbance during
Thresholding value is depends upon the sampling real time ECG signal attain. Now features extract of
points , time in sec of original ECG signal, technique signal with the help of array create on front panel.
is applied directly by right click on wavelet type
options are available [11].

Figure10. Simulation model of ECG signal analysis.

Wavelet concept: Wavelet are being useful in  Wavelet separates a signal into multi-
various fields of science and engineering such as resolution components [14] .
signal and medical imaging processing , signal Results: The simulation result and performance of
processing etc. The wavelet transform gives a large ECG parameters and classification was trained in
numbers of small coefficients and a small number of combination of wavelet transform with soft threshold
large coefficients. Small coefficients mainly represent on 8 level wavelet coefficients . The sampling
the noise components and large coefficients represent frequency of the ECG signal is 360 Hz. Figure 11
the signal values of raw signal. Following shows the raw of ECG signal which is contaminated
characteristic which make them useful are given: with noise , figure 12 tells us baseline wandering is
 Wavelet are localized in time and frequency removed with De-trend of Wavelet to smooth the
domain. signal from any noise , figure 13 explains power line
 For analysis non-stationary signal such as interference of noisy ECG signal with Wavelet de-
ECG, EEG etc. noise Express VI waveform to get the proper
diagnosis.

Figure11 Raw ECG signal infected with noise

Parameters Calculation and Classification of ECG Signals Using Lab View

66
International Journal of Industrial Electronics and Electrical Engineering, ISSN: 2347-6982 Volume-4, Issue-9, Sep.-2016

Figure 12 Baseline wandering removed by Wavelet Denoise VI

Figure13 Simulation results with maximum de-noise by UWT

CONCLUSION Diagnosis of Heart Diseases”,Sensors and Transducer

Journal,vol.133, Issue 10, October 2011.
[5] DL Donoho, "De-Noising by Soft-Thresholding," IEEE
In this paper, a Simulation model of ECG signal are Transactions on Information Theory, Vol. 41, pp. 613-27,
often corrupted by power line interference and 1995.
baseline wandering noise that need to be removal [6] Khalifa Elmansouri, R. Latif, B. Nassiri and S.
Elouaham,” New Electrocardiogram Signal Analysis in a
before processed of ECG signal can be used to doctor
Research Laboratory Using LabVIEW”, International
for analysis of disease. The processed system are of Journal of Interdisciplinary Research and Innovations
two types: features extraction phase and classification (IJIRI), Vol. 1, Issue 1, pp: (15-21), Month: October-
phase. In the first phase, moving average filters is December 2013
employed to eliminate the baseline noise from the [7] P.E. Tikkanen, “Nonlinear wavelet and wavelet packet
denoising of electrocardiogram signal,” Biol. Cybern.
ECG signals. Then the UWT is applied on filtered Vol.80, 1999, pp.259-267.
signal and some features from wavelet are extracted. [8] R. R. Coifman and Y. Meyer, “Orthonormal wave packet
The simulation results demonstrated the processed bases,” Math. Dept., Yale Univ., New Haven, CT, Tech.
system could be employed for the classification and Rep., 1990.
[9] S. Mahmoodabadi, A. Ahmadian, and M. Abolhasani,
parameters of noised signal. "ECG feature extraction using Daubechies wavelets," in
Proceedings of the fifth IASTED International
REFERENCES [10] Goutam Kumar Sahoo, Samit Ari,Sarat Kumar Patra,”
ECG signal analysis for detection of Heart Rate and
Ischemic Episodes” International Journal of Advanced
[1] Rangayyan R.M., Biomedical Signal Analysis, Wiley-
Computer Research (ISSN (print): 2249-7277 ISSN
Intersciene, John Wiley & SONS, INC., 2002.
(online): 2277-7970) Volume-3 Number-1 Issue-8 March-
[2] Lina Zhang, Xinhua Jiang, “Acquisition and Analysis
2013.
System of the ECG Signal Based on LabVIEW”, The
[11] Akshu Purohit, Khalid Khan, Govind Kishan Bohra,”
Ninth International Conference on Electronic
Calculate ECG Parameters through Labview”, National
Measurement & Instruments ICEMI’2009.
Conference on Innovations and Recent Trends in
[3] Donghui Zhang, “Wavelet Approach for ECG Baseline
Engineering and Technology (NCIRET-2014)
Wander Correction and Noise Reduction”, Proceedings of
[12] National instrument LABVIEW ECG signal processing.
the 2005 IEEE Engineering in Medicine and Biology 27th
[13] www.physionet.org
Annual Conference Shanghai, China, September 1-4,
[14] Nagendra H, S.Mukherjee, Vinod Kumar” Application of
2005.
wavelet techniques in ECG signal processing”, IJEST
[4] Channappa Bhyri, Satish T Hamde, Laxman M
journal of research,vol.3 no.10 October 2011.
Waghmare,”ECG Acquisition and Analysis System for



Parameters Calculation and Classification of ECG Signals Using Lab View

67