Download The ECG data compression performance is presented

International Journal of Electrical, Electronics and Computer Systems, (IJEECS)
_______________________________________________________________________
ECG DATA COMPRESSION USING PRINCIPAL COMPONENT
ANALYSIS
1
Seema Deshpande, 2S.O.Rajankar
1,2
Sinhgad College of Engineering, Pune , India.
Email : 1seemadeshpande293@gmail, [email protected]
compression methods can classified into two main
families: lossless and lossy methods. Methods from the
lossless can obtain an exact reconstruction of the
original signal, but they do not get low data rates. In
contrast, lossy methods do not obtain an exact
reconstruction, but higher compression ratios can
obtained. The commonly used ECG compression
techniques are lossy in nature [3]
Abstract- ECG signal analysis has shown an important role
in the diagnosis of heart diseases. ECG signal compression
is required due to three main reasons: low storage data
space, reduction of low data transmission rate and
transmission
bandwidth
conversation.
The
electrocardiogram (ECG) signal compression using
principal component analysis (PCA) is presented in this
paper. Principal Component Analysis (PCA) technique is
used for dimensionality reduction and data classification.
The methods are applied to the MIT/BIH arrhythmia ECG
database. we are going to work with the first minute of an
ECG sample taken from file 106 of the MIT-BIH
Arrhythmia Data base. The results are efficient for data
compression of ECG signals. The experimental results are
analyzed on the basis of compression ratio (CR).
Keyword: - Principal
Compression ratio.
component
Analysis
II. ELECTROCARDIOGRAM SIGNALS
An Electrocardiogram is also called an EKG or ECG.
Electrocardiography (ECG or EKG) is an electrical
activity of the heart over a period of time, as detected
electrodes attached to the surface of the skin and
recorded by a device external to the body. The recording
produced by this noninvasive procedure is termed an
electrocardiogram (ECG or EKG). An ECG is used to
measure the regularity of heartbeats, as well as the
position and size of the chambers, the presence of any
injure to the heart, and the effects of devices used to
regulate the heart, such as a pacemaker [2]. Most ECGs
are performed for diagnostic or research purposes on
human heart, but may also be used performed on
animals, usually for prognosis of heart abnormalities or
research. Figure below show normal human
electrocardiogram waveform [2].
(PCA),
I. INTRODUCTION
The ECG (Electrocardiogram) is a biological signal. It is
the electrical activity of the heart. The main purpose of
compression is to represent an ECG data with the
smallest possible number of bits. Many types of ECG
recordings generate a vast amount of data. While ECG
system found primarily in hospitals, they find use in
many other medical centre’s also. A growing area of use
for ECG is 24-hour continuous ECG performed in
intensive care units and stress test ECG [1]. With the
growing use of these ECG signals to detect and diagnose
heart disorders, by compressing the ECG data more
information can be stored and processed for future
evaluation.
The main goal of any compression technique is to
achieve maximum data reduction and preserving
morphology
features
upon
reconstruction.Data
__________________________________________________________________________
ISSN (Online): 2347-2820, Volume -1, Issue-2, 2013
13
International Journal of Electrical, Electronics and Computer Systems, (IJEECS)
_______________________________________________________________________
component x1 is a linear combination of (y1,y2,……yp),
that is
x1 = (a11y1+a12y2+………..+a1pyp
x1=(a11y1+a12y2+.................a1pyp)
p
x 1 = ∑j=1 a1j yj=aT1 y
The principal component x1 is such that its variance is
maximized given the constraint that a 1Ta1= 1. Principal
component analysis finds the optimal weight vector
(a11,a12,…….a1p) and associated variance of x1 which is
usually denoted as λ1. The second principal component
involves
finding
a
second
weight
vector
(a21,a22,………,a2p) such that the variance x2 is
maximized such that a2Ta2 = 1 and associated value is
denoted by λ2.This process can be continued until as
many component as variables have been calculated. The
sum of variance of principal component is equal to the
sum of the variance of original variables such that
Fig 1. shows the basic ECG waveform
III. ECG DATA COMPRESSION
There are two types of data compression technique
namely
1.
lossless compression: Lossless data compression is
that allow the exact original data to be
reconstructed form the compressed data.
2.
lossy compression: In lossless compression
technique original data after compression and
decompression are found exactly same form.
Lossless method is used when we cannot afford to
lose any data legal and medical document
computer program [1]. Lossy compression
provides much higher compression ratios than
lossless. Figure 2 shows the block diagram of
lossless compression of signal.
(1)
∑pj=1 λj = ∑pj=1 σ2j
p
(2)
j
x 1 = ∑j=1 a1j yj………xp=a p yj
(3)
Principal
component
analysis
by
eigenvalue
decomposition finds the directions in the data with the
most variation that is the eigenvector corresponding to
the highest eigenvalue of the covariance matrix, and
projection the data onto this direction. The Principal
component analysis transformed by data is as
X=V T Y
(4)
Where V is the eigenvector. The eigen vector on the
bases of highest eigen value.
Fig 2.Lossless compression
V. STEPS OF ECG DATA COMPRESSION
USING PCA
IV. PRINCIPAL COMPONENT ANALYSIS
The idea of principal component analysis (PCA) is to
reduce the dimensionality of a data set. This is get by
transforming to a new set of variables, the principal
components (PCs), which are uncorrelated, and which
are ordered so that the first few retain most of the
variation present in all of the original variables. PCA is
concerned with explaining the variance-covariance
structure of a set of variables through linear
combinations of these variables. Consider a random
vector Y= (y1, y2,…….…..,yp) the covariance matrix of
Y is C and eigen values are λ1, λ2, λ3…..... λp such that
λ1 ≥λ2≥……….≥ λp and the eigenvector and eigen values
pairs are (V1,λ1),(V2, λ2)…(Vp, λp). The principal
components (PCs) are linear combination (x1,x2,…xp)
with the changed coordinate system of p random
variables (y1, y2,………Yp). The first principal
The following simplified steps are applied to the PCA
based ECG compression analysis:
Step 1. Get ECG data MIT-BIH Arrhythmia data base).
Step 2. Calculate the mean of ECG data.
Step 3. Subtract the mean from ECG data.
Step 4. Calculate the covariance matrix of subtract data.
Step5. Calculate eigenvectors and Eigen values of the
covariance matrix.
Step 6. Sort the Eigen values and Eigen vectors, on the
basis of highest Eigen values.
Step 7. Principal components choose and form ECG
feature vectors = [eig1;eig2;eig3;……eigL]
__________________________________________________________________________
ISSN (Online): 2347-2820, Volume -1, Issue-2, 2013
14
International Journal of Electrical, Electronics and Computer Systems, (IJEECS)
_______________________________________________________________________
Step 8. Deriving the new ECG data i.e.
New data =[Row Feature Vector]T*[Row data adjust]
Step 9. Calculate the compression ratio (CR).
VI. MEASURE OF PERFORMANCE
One of the most difficult problems in ECG compression
applications and reconstruction is defining the error. The
main purpose of the compression system is to remove
redundancy and irrelevant information. The ability of
the reconstructed signal to preserve the relevant
information. ECG signals generally are compressed with
lossy compression algorithms, a way of quantifying the
difference between the original and the reconstructed
signal, often called distortion [4].
A. Compression ratio
Fig 4.The reconstructed compressed of ECG data
Compression efficiency is measured by the compression
ratio. The compression ratio (CR) is defined as the ratio
of the number of bits representing the original signal to
the number of bits required to store the compressed
signal [6]-
The output signal is one fourth length that of its original
signal [7].
The ECG data compression performance is presented
CR. ECG data generally are compressed with lossy
compression algorithms. Data compression algorithm,
used to minimize data storage by eliminating the
redundancy wherever possible to increase the
compression ratio. Compressed data must also represent
high compression ratio.
Compression ratio of
reconstructed ECG is 2.2. The various performance
evaluation parameters like CR obtain in MATLAB
R2009a using principal component analysis (PCA). A
data compression algorithm must represent the data with
acceptable fidelity while archiving high compression
ratio (CR).
VII. RESULTS
VIII. CONCLUSION
The experimental data from MIT-BIH arrhythmia
database is used to analyze and test the performance CR.
Figure (3) and bellow shows original ECG waveform.
The reconstructed compressed of ECG data shown in
figure (4).
In this paper, ECG data compression proposed scheme
based on principal component analysis (PCA) algorithm.
The effectiveness and performance of an ECG data
compression evaluated compression ratio (CR). The
parameters most commonly used by the international
community like percent root mean square difference
(PRD) and compression ratio (CR). If ECG data
compression high-compression ratio (CR) and i.e. ECG
data good compressed.
Compression ratio =
N∗12
N
∗Number of bits required
4
(6)
Where N is the number sample of data.
Number bits required=log2(abs(max(output)))+1
REFERENCES
Fig 3.original ECG waveform
[1]
Hemlata Shakya, A. K.Wadhwani,“Transform
based ECG Data Compression”, International
Journal of Engineering and Advanced
Technology (IJEAT) ISSN: 2249-8989,
Volume1, issue-4, April2012.
[2]
http://en.wikipedia.org/wiki/Electrocardiograph
y.
[3]
Vibha Aggarwal and Manjeet Singh Patterh,
“ECG Compression using wavelet packet,
Cosine Packet and wave Atom transforms”, ,
International
Journal
of
Electronics
Engineering research ISSN 0975 Volume 1
number 3(2009 )pp.259-268 © Research India
publications http://ripublication/ijeer.htm.
__________________________________________________________________________
ISSN (Online): 2347-2820, Volume -1, Issue-2, 2013
15
International Journal of Electrical, Electronics and Computer Systems, (IJEECS)
_______________________________________________________________________
[4]
Ruquiya khanam, Syed Naseem Ahmad,
“Selection of wavelets for Evaluation,
SNR,PRD and CR of ECG signal”,
International Journal of Engineering Science
and Innovative Technology (IJESIT) Volume 2,
Issue 1, January 2013
Algorithm for Classification of Compressed
ECG”, I.J. Information Technology and
Computer Science, 2012, 2, 26-33, Published
Online
March
2012
in
MECS
(http://www.mecs-press.org/),
DOI:
10.5815/ijitcs.2012.02.04.
[5]
Er. Abdul Sayed, “ECG DATA Compression
using DWT & HYBRID”, International Journal
of Engineering Research and Application
(IJERA), ISSN: 2248-9622, vol. 3, Issue 1,
January-February 2013, pp.422-425.
[7]
Leslie Cromwell, Fred J. Weibell, Erich A.
Pfeiffer, Biomedical Instrumentation and
Measurement, Second Edition, Prentice-Hall of
India Private Limited New Delhi-110001,
2006.
[6]
Shubhada S. Ardhapurkar, Ramandra R.
Manthalkar, Suhas S. Gajre, “A Hybrid
[8]
MIT-BIH
arrhythmia
database
http://www.physionet.org /cgi-bin/atm/ATM
__________________________________________________________________________
ISSN (Online): 2347-2820, Volume -1, Issue-2, 2013
16