International Journal of Pure and Applied Mathematics

Volume 119 No. 14 2018, 79-85

ISSN: 1314-3395 (on-line version)
url: http://www.ijpam.eu
Special Issue
ijpam.eu

An Efficient Real Time Arrhythmia Detector Model using

LABVIEW
Renuka V Tali[1], Pratibha S N[2], Gunasheela T J[3], Shilpa A P[4]
[1,2,3,4]Assistant Professor, Department of ECE, KSSEM, Bangalore, Karnataka, India

Abstract

Human research and knowledge even after such advancement in the technology, available information is yet not capable of reducing or
removing cardiac disorder. Hence there is a need for introduction of an instrument which can detect Arrhythmia and can help doctors to
treat patients in a more improvised way. To develop such a device, we can use LABVIEW software which can help in real time testing of
the ECG signals with great speed and accuracy. Therefore, the purpose of our project is to construct an arrhythmia detector that can
detect and identify basic types of arrhythmia. Our model basically has two parts, hardware and software. Hardware part consists of 3
surface electrodes, instrumentation amplifier and filters. ECG is extracted from human subject using surface electrodes. Since these bio
signals are weak in amplitude, they are passed through instrumentation amplifier followed by a low pass filter and then by a high pass
filter and that is sent to LabVIEW software through an USB DAQ for further analysis. Software part includes analysis in LabVIEW
using biomedical workbench. Feature extraction is done in LabVIEW software using wavelet transforms and peak detection algorithm.
BPM ranges are set according to three cases like Normal, Tachycardia and Bradycardia. If arrhythmia is detected then through GSM
module, the message of arrhythmia detection is sent to doctor in the text form.

Key words: USBDAQ, ECG, BPM, Wavelet Transform, LabVIEW.

Acquisition Card the data can further be processed using

LabVIEW software and the desired calculations can be performed
1. Introduction and occurrence of basic arrhythmia can be detected.
Arrhythmia is a case in which heart loses its rhythm of Cardiac arrhythmias are one of the leading causes of
synchronous beating, this can be both painful and at times lethal cardiovascular disease(CD) for men and women. Diagnosis and
too. Hence there is a need of introduction to an instrument which treatment of these conditions are usually achieved using
can detect Arrhythmia and can help doctors to treat patients in a electrocardiogram (ECG) devices such as holter monitors, loop
more improvised way. For developing such detectors, we can use recorders and Implantable Cardioverter Defibrillators(ICD) among
software products which can help in real time testing of the ECG others. In this document, we report the development of a simple,
signals with great speeds and accuracy. This bodes well for such a single lead, real time ECG device intended for the monitor and
device being developed with significantly better accuracy and detection of ventricular tachycardia(VT). Control software was
arrhythmia detecting capabilities and can further be linked with developed using LabVIEW’s biomedical toolkit and validated
artificial intelligence software so as to detect any further using simulated ECG signals [3].
occurrence of arrhythmias on the basis of the prior collected data
patterns[1].
2. Literature Survey
Preliminary detection and classification of cardiac Arrhythmia is
one of the most important problems in biomedical signal In [4] for extracting parameters of ECG we use wavelet transform,
processing. Computer assisted Arrhythmia recognition is critical wavelet analysis breaks a signal down into its constituent parts for
for the management of Cardiac disorders. Arrhythmia is very analysis. Mother wavelet transform that we use here is Daubechies
common cardiovascular disease and there are more than 10 of D6 family. If a signal is not well represented by one member of
million cases per year [2]. the Db family, it may still be efficiently represented by another.
Daubechies wavelet family has similar shape to QRS complex and
Arrhythmia is a life threatening situation that requires immediate their energy spectrum is concentrated around low frequencies The
medical assistance. Continuous monitoring of patient is required, wavelet transform is a convolution of the wavelet function ψ (t)
as there are no proper symptoms. Hence there is a need to develop with the signal x (t).
of an instrument which can detect Arrhythmia and can help
doctors to treat patients in a more improvised way. Orthonormal dyadic discrete wavelets are associated with scaling
functions ϕ (t) [5].
The goal of this project is to develop a device that would replace
the existing Electrocardiograms and facilitate doctors to record In [6], many linear techniques for VT/VF detection have been
Electrocardiographs and detect basic arrhythmias simultaneously. developed, such as the probability density function method, rate
With advent of technology and use of software like LabVIEW a and irregularity analysis, analysis of peaks in the short-term
virtual instrument can be designed which can be programmed by autocorrelation function, sequential hypothesis testing algorithm,
the users. Hence while developing Arrhythmia detector we choose correlation waveform analysis, four fast template matching
to develop a hardware which would acquire the ECG signals and algorithms, VF-filter method, spectral analysis, and time-
further amplify and filter it and then by means of a Data frequency analysis. However, these methods exhibit

79
International Journal of Pure and Applied Mathematics Special Issue

disadvantages, some being too difficult to implement and 3. Design And Implementation
Compute for automated external defibrillators (AED’s) and

Implantable cardioverter defibrillators (ICD's), and some only 3.1 Block Diagram
successful in limited cases. For example, the linear techniques
using the features of amplitude or frequency have shown their
limits, since the amplitude of ECG signal decreases as the VF
duration increases, and the frequency distribution changes with
prolonged VF duration. Therefore, more sophisticated signal
processing techniques are needed to fully describe and
characterize VT and VF and facilitate the development of new
detection schemes with high correct detection rate, or
equivalently, with low false-positive and false-negative
performance statistics [6].

Curve Length Transform(CLT) is used for QRS detection and

Discrete Wavelet Transform(DWT) is utilized for T and P wave
detection. CLT offers a computationally efficient QRS detection
technique. The CLT is pipelined and requires only one square, one
square root and summation for each sample. And the DWT is
implemented as cascade of filters [7].

The estimation of instantaneous frequency (IF) of an ECG signal Fig. 1 Block Diagram of Arrhythmia Detector
is used as a method for carrying out detection of cardiac disorder. ECG signals are extracted from the human body with the help of
Based on IF estimates, a classifier has been designed to surface electrodes [15], the signals are passed through the
differentiate a diseased signal from a normal one. Training, testing instrumentation amplifier to reduce the noise. Then it is passed
and validation of the classifier has been carried out using signals through the Low pass and High pass filter respectively. Now in
from MIT Arrhythmia, normal Sinus Rhythm and Ventricular order to get high gain the signal is passed through Non inverting
Arrhythmia databases, respectively [8]. The sensitivity and amplifier. The output of Non-inverting amplifier is given to NI
specificity of the classifier comes out to be 97.82% and 100%, myDAQ [17]. DAQ is the interface between hardware and
respectively and it is considered especially suitable for ambulatory LabVIEW software. In LabVIEW feature extraction is made with
ECG analysis. The method normalizes the difference in the ECG the help of wavelet transformation and The arrhythmia detection is
morphology among subjects using dedicated wavelets. To evaluate made with the help of peak detection algorithm. If Arrhythmia is
the effectiveness of the normalization process, we compared five detected, then the output will be sent as a text message to the
feature extraction methods. Those are a normal template cross- Doctor using GSM module.
correlation (Template M), a discrete wavelet transform (DWT)
with the Haar wavelet (Haar (7)), a DWT with the Daubechies 6
3.2 Pcb Circuit And Layout
wavelet (db6(4)), a CWT with the dedicated wavelet of subject
106 (106CWT), and a CWT with the dedicated wavelet of each
subject [9]. Each feature extraction method was applied to the
ECG signals of the N class and V class, which were extracted
from subjects who had over 100 heartbeats of the V class. Using
the methods above, we compressed the output data to one-
dimensional data via LDA. For 106CWT and Dedicated, we used
the PCA-LDA composition, because these two methods are based
on CWT. The compressed data was normalized to remove
amplitude variation due to the different signal processing methods
[14].

Arrhythmia detection and classification is by estimating IF of a

Fig. 2 PCB Circuit and Layout
single lead ECG waveform, rather than the usual RR interval time
series. IF has been estimated by utilizing the method of average The hardware circuit consists of amplifiers and filters. Supply
spectral density. The output of the IF estimator is utilized to voltage for all the filters and amplifiers is given by a 12V battery.
design a classifier which differentiates a diseased signal from a It ensures the isolation of the circuit from the short circuit.
normal one [15].
 Amplifier.
In [16] is to present a real-time detection method for cardiac event  Filter.
detection using the intra cardiac EGM signals. Sensed EGM
signals differ from ECG signals in many aspects. Major Amplifiers: The signal that is extracted from the wrist is of low
differences are summarized here. EGM signals provide direct amplitude, so the signal needs to be strengthened. We use INA
access to individual heart chambers, most importantly right 217 at the starting to amplify this signal. To amplify the output
ventricular and right atrial, at the signal source. In contrast, ECG from the high pass filter we use a noninverting amplifier, which
signals provide a combined signal after propagation of various amplifies the signal so that it can be sent into DAQ.
waves to the body surface. Relative timing of various
EGM signals is very important as it could be employed to Filters: Filters are circuits which perform signal processing
discriminate various cardiac events accurately for ECG signals, functions, specifically to remove unwanted frequency components
however, such timing information is not available [16]. from the signal, to enhance wanted ones, or both. The filters used
are Low pass filter (LPF) and High pass filter (HPF) [7, 21].

80
International Journal of Pure and Applied Mathematics Special Issue

3.3 NI myDAQ 3.5 Gsm Module

Fig. 5 GSM module

If arrhythmia is detected, then notification should be sent to doctor

via GSM module. An alarming about abnormal functioning of
heart can help doctors take appropriate measures at right time.
Here we use SIM800 GSM module to communicate

3.6 Arduino Uno

Fig. 3 NI myDAQ

It is affordable for student use and powerful enough for more

sophisticated measurement applications [11,20]. Mac OS X and
Linux users can download the NI-DAQmx Base driver software
and program the USB6009 with LabVIEW. To supplement
simulation, measurement, and automation theory courses with
practical experiments, NI developed a USB-6009 Student Kit that
includes a copy of LabVIEW Student Edition. It is used to
interface hardware and software.

3.4 Labview
Fig. 6 Arduino uno

In this project we use Arduino uno to control the GSM module we

can enter the mobile number of the doctor and the message to be
sent when the arrhythmia is detected in the Arduino software
program. The Arduino UNO is a widely used open-source
microcontroller board based on

the ATmega328P microcontroller and developed by Arduino.cc.

The board is equipped with sets of digital and analog input/output
(I/O) pins that may be interfaced to various expansion boards
(shields) and other circuits [10].

4. Results And Discussions

Fig. 4 Window Block Diagram
The front panel of LabVIEW consists of raw ECG graphs, peak
Software used for processing the signal by programming is plot graph, BPM, instantaneous heart rate, number of peaks. There
LabVIEW. LabVIEW software is used to remove additional noise is a light notification for all three conditions i.e for normal,
from the ECG signal and to detect arrhythmia. We make use of tachycardia, bradycardia. There is another light notification for
two panels or windows, one is the front panel and second one is arrhythmia detection, which turns red when tachycardia or
the block diagram window. In front panel window the ECG bradycardia is detected and remains green when normal ECG.
waveform, peak plot, features such as P onset, P offset, QRS, T
onset and offset, ST level, iso level, instantaneous heart rate, On the right corner there is a array of features like P onset, P
number of peaks [12,13] and arrhythmia detection display can be offset, QRS onset, R, QRS offset, T onset, T offset, amplitude, iso
viewed. Feature extraction settings such as detrend settings can be level, ST level, P wave, T wave.
changed. The block diagram window consists of blocks or
graphical programs for arrhythmia detection and other features The output of amplifier and filter circuit is fedto LabVIEW
through DAQ. The signal goes through wavelet denoise block,
wavelet analysis and feature extraction block. The detrend settings
is used to remove the baseline wandering. Three types of inputs
are used for analysis in LabVIEW, first one is through a
simulation block, second one through DAQ assistant block and
other one through read from measurement file block. Simulation
block is used to give simulated signals, the DAQ assistant block

 
 
 

81
International Journal of Pure and Applied Mathematics Special Issue

can be used to give real time signal from the subject and read from
measurement file block is used to give recorded signals or
database signals(MIT-BIH) [18,19,21].

Here the BPM (beats per minute) is 40, hence the bradycardia
condition button becomes light green from dark green. The
arrhythmia detected button turns on red.

Fig. 10 text alert to the doctor

The doctor takes appropriate measures to treat the patient after

receiving this message. The doctor need not be present near the
patients all the time.

Table of Test Results for Simulated Signal.

Input signal Expected Detected Condition

Fig. 7 Output for normal ECG signal
output output detected
Here the BPM (beats per minute) is 80, hence the normal
condition button becomes light green from dark green. The Normal Arrhythmia Arrhythmia Normal
arrhythmia detected button remains off in dark green. not detected not detected

Tachycardia Arrhythmia Arrhythmia Tachycardia

detected detected

Bradycardia Arrhythmia Arrhythmia Bradycardia

detected detected

Table shows the test results for simulated signal. The detected
output is same as that of expected output. The accuracy is 100% in
this case.

Table of Testing Results for Real-Time DAQ Input.

Fig. 8 Output for Tachycardia
Patient Patient BPM Expected Detected
Here the BPM (beats per minute) is 240, hence the tachycardia
condition button becomes light green from dark green. The name condition output output
arrhythmia detected button turns on red. shabarish Normal 75 Arrhythmia Arrhythmia
not detected not detected
shwetha Normal 85 Arrhythmia Arrhythmia
not detected not detected
charu Normal 83 Arrhythmia Arrhythmia
not detected not detected
ramya Normal 80 Arrhythmia Arrhythmia
not detected not detected

Table gives the testing results for real-time DAQ input. We tested
on 4 patients, all are in normal condition so expected output is
“arrhythmia not detected”. The detected output is same as
expected output in all four cases, therefore accuracy is 100%.

Fig. 9 Output for Bradycardia Accuracy = 4/4 *100 =100%

Table of Test Result for MIT-BIH Database [18,19].

Input Existing Detected Detected

condition condition output
Normal 1 Normal Normal Arrhythmia
not detected

82
International Journal of Pure and Applied Mathematics Special Issue

Normal 2 Normal Normal Arrhythmia Table shows the test result for MIT-BIH database. We tested on
not detected 30 patients, out of these only 28 are detected correctly. The
Normal 3 Normal Normal Arrhythmia expected condition is not same as detected condition for two
not detected cases, therefore the accuracy is 93.33%.
Normal 4 Normal Normal Arrhythmia
Accuracy=28/30 * 100= 93.33%
not detected
Normal 5 Normal Tachycardia Arrhythmia Sensitivity = 19+119 *100= 95%
detected
Normal 6 Normal Normal Arrhythmia Specificity 9/(9+1) * 100= 90%
not detected
Table of Comparison of Accuracy
Normal 7 Normal Normal Arrhythmia
not detected
Author Database No of Parameter Accuracy
Normal 8 Normal Normal Arrhythmia
samples considered obtained
not detected
P. MIT-BIH 19 QRS 86.61%
Normal 9 Normal Normal Arrhythmia Keerthi Complex
not detected priya,
Normal 10 Normal Normal Arrhythmia 2015
not detected Sanket MIT-BIH 3(100,101, QRS Not
Tachycardia Tachycardia Tachycardia Arrhythmia Rege, 200) Complex mentione
1 detected 2015 d
Ping MIT-BIH 50 QRS 92.69%
Tachycardia Tachycardia Tachycardia Arrhythmia
cheng, Complex
2 detected
2017
Tachycardia Tachycardia Tachycardia Arrhythmia Proposed MIT-BIH 30 BPM 93.33%
3 detected Project
Tachycardia Tachycardia Tachycardia Arrhythmia
4 detected
4. Conclusion
Tachycardia Tachycardia Arrhythmia
5 Tachycardia detected
This project can provide doctors with better information related to
Tachycardia Tachycardia Tachycardia Arrhythmia patient’s heart conditions and help them to identify further
6 detected chances of occurrences of arrhythmia. In future it can be modified
Tachycardia Tachycardia Arrhythmia to detect other types of arrhythmias. The software part i.e
7 Tachycardia not LabVIEW virtual instrument block window and the front panel
detected can be used with even other ECG extraction hardware. The major
Tachycardia Tachycardia Tachycardia Arrhythmia
advantage of using the methods used for arrhythmia detection in
this model is that the parameters of the code can be easily changed
8 detected
and customized to an individual's unique heart waveform.
Tachycardia Tachycardia Tachycardia Arrhythmia
9 detected An individual's waveform may change over the course of years, or
Tachycardia Tachycardia Tachycardia Arrhythmia the waveform may vary from person to person with respect to his
10 detected or her age and condition of heart and his or her health and fitness.
Bradycardia Bradycardia Bradycardia Arrhythmia By adjusting the parameters which determine whether an event
1 detected has taken place in the subject's unique waveform, either a beat or
particular feature of an arrhythmia, the code will detect fewer false
Bradycardia Bradycardia Bradycardia Arrhythmia
parameters. The code will detect the features of the arrhythmias
2 detected
with very good accuracy; however, it is still may not be correct all
Bradycardia Bradycardia Bradycardia Arrhythmia of the time. An improvement would be to increase the accuracy of
3 detected the code.
Bradycardia Bradycardia Bradycardia Arrhythmia
4 detected Model has been tested for three different input sources, real time
Bradycardia Bradycardia Bradycardia Arrhythmia input, simulated input and input from MIT-BIH data base. The
5 detected
model proves 100% accurate for real timeinput and simulated
input. 93.33% of accuracy is obtained for 30 samples from
Bradycardia Bradycardia Normal Arrhythmia
database collected from different individuals. Performance metric
6 not detected such as sensitivity and specificity are also calculated and found to
Bradycardia Bradycardia Bradycardia Arrhythmia be about 95% and 90% respectively. Results obtained from
7 detected implemented model is compared with other published models and
Bradycardia Bradycardia Bradycardia Arrhythmia proves better for parameters considered.
8 detected
Bradycardia Bradycardia Bradycardia Arrhythmia 5. Future Work
9 Detected
Bradycardia Bradycardia Bradycardia Arrhythmia Making the device portable: By using wireless electrodes which
10 detected can transfer the received ECG signals via GPS or Bluetooth or any
 
 
 

83
International Journal of Pure and Applied Mathematics Special Issue

such wireless transmission software, to a remote computer or [21] C J finelli, “The Time Sequenced Adaptive Algorithm: Application
Laptop having the Arrhythmia Detector so that it can be to Morphological Adaptation and Arrhythmia Onset Detection”,
continuously monitored. Can include detection of more types of IEEE, 1992.
Arrhythmias: Currently the Project module is just detecting a
selected few types of Arrhythmias which occur on regular basis.
In future other arrhythmias like PVC, flutters, Atrial fibrillation,
ventricular Fibrillation can also be introduced for detection. To
interface the equipment with artificial intelligence soft-wares to
detect any further occurrence of such arrhythmias. This can be
achieved by comparing existing simulated patterns with the patient
ECG and by doing Template matching so as to detect any similar
patterns before an occurrence of an arrhythmia.

Acknowledgement: We sincerely acknowledge the support and

assistance of our final year graduating students in carrying out this
work.

References
[1] Shilpa Khandke, Jyoti Warrier, Dr.C.D.kapse, “Arrhythmia
Detector Using LABVIEW”, IJERT, volume 02, issue 04, pages 05,
April 2013.
[2] Sanket Rege, Tyler Barkey, Mariano Lowenstern, “Heart
Arrhythmia Detection”, IEEE, pages 07, 2015.
[3] Akib Mohammad Azam Zaidi, Muhammad Jubaer Ahmed,
A.S.M.Bakibillah, “Feature Extraction and Characterization of
Cardiovascular Arrhythmia and Normal Sinus Rhythm from ECG
signals using LABVIEW”, IEEE, pages 6, 2017.
[4] P.KeerthiPriya,Dr.G.UmamaheswaraReddy, “MATLAB based GUI
for Arrhythmia Detection using Wavelet Transform”, IJAREEIE,
volume 04, issue 08, pages 10, February 2015.
[5] Junjiang Zhu, Lingsong He and Zhiqiang Gao, “Feature Extraction
from a Novel ECG Signal Model for Arrhythmia Diagnosis”, bio-
medical materials and engineering, 2014.
[6] Ping chen, Xiaodai dong, “Life Threatening Ventricular Arrhythmia
Detection with Personalised Features”, 2017.
[7] Temesghen Tekeste, Hani Saleh, Baker Mohammad, Hasan
Kandokar, Mohammaad Ismail, “A Biomedical Soc Architecture
for Predicting Ventricular Arrhythmia”, IEEE, 2016.
[8] Paul Couto, Ruben Ramalho, Rui Rodrigues, “Suppression of False
Arrhythmia Alarms Using ECG And Pulsaticle Waveforms”,
computing in cardiology, 2015.
[9] Keun her, Chi bum aho, Sung min park, Seong wook choi, “Heart
Monitoring Using Left Ventricular Impedence And Ventricular
Electrocardiography Inn The Left Ventricular Assist Device
Pateints”, 2015
[10] M.Sabarimalai Manikandan, “ Straightforward Robust QRS
Detection Algorithm For Wearable Cardiac Monitor”, healthcare
technology letters , 2014.
[11] Taegyun jeon, “Implementation of A Portable Device Real-time
ECG Signal Analysis”, Biomedical Engineering, 2014.
[12] Farid rezazadeh, “A New Heart Arrhythmia, Detection Algorithm”,
2013.
[13] Saed karimifard and Alireza ahmadian, “A Robust Method for
Diagnosis of Morphological Arrhythmia Based On Hermitian
Model of Higher Order Statistics”, 2011.
[14] Jinkwan kim, “An Arrhythmia Classification Algorithm using a
Dedicated Wavelet Adapted to Different Subjects”, biomedical
engineering, 2011.
[15] Asin Dilawer Bakshi and Muhammad Ali, “Cardiac Arrhythmia
Detection using Instantaneous Frequency Estimation of ECG
Signals”, IEEE, 2010.
[16] Hamid Sheikhzadesh, Robert L Brennan, “Real-time Cardiac
Arrhythmia Detection using Wola Filter Bank Analysis of EGN
Signals”, 2007.
[17] Yan Sun, Kap Luke Chang and Shankar Muthu Krishnan, “Life
Threatening Ventricular Arrhythmia Recognisation by Non Linear
Discripter”, biomedical engineering, 2005.
[18] Ivaylo I Christov, “Real Time Electrocardiogram and QRS
Detection using Combined Adaptive Threshold”, springer, 2004.
[19] Dingfei Ge, narayanan Srinivasan and Shankar M Krishnan,
“Cardiac Arrhythmia Classification using Autoregressive
Modelling”, biomedical engineering, 2002.
[20] Kang-pinglin And Walter H.Chang, “A Technique for Automated
Arrhythmia Detection of Halter ECG”, IEEE, 1995.

84
85
86