ECG Data Compression: A Review
Vineet Kumar1, B. S. Saini2, R. K. Sunkaria3
1 23
Electronics and Communication Department, Dr. B. R. Ambedkar National Institute of Technology,
Jalandhar
1
vineetbcrec@gmail.com, 2 sainibss@gmail.com, 3 sunkariark@gmail.com
Abstract-The recent development of biomedical signal
processing, information technology and
communication has brought a new dimension to the
medical world. It is now possible to record an ECG
signal with a portable electrocardiograph lightweight
and able to transform a micro-computer in
electrocardiograph with the possibility of a diagnostic
aid for automated interpretation. Applications
focusing on portable devices for 24-hour on-line
cardiac monitoring are in increasing demand. We are
now shifting from a paper-based patient record to a
digital one. Many ECG compression algorithms have
been researched for optimization with respect to
storage capacity and network bandwidth. For this we
have to achieve a high compression ratio (CR) result
and preserve clinical information (low PRD).
Examples of ECG compression algorithms include
AZTEC , TP, CORTES, KLT , DCT, DWT .Among
these, wavelet transform-based algorithms have
received a great deal of attention over the past several
years because of their straightforward
implementation and efficiency, and their good
localization properties in time and frequency
domains. This paper gives a bibliographical survey
and general backgrounds of research and
development in the field of ECG data compression
based on over 50 published articles. The collected
literature has been divided into many sections so that
new researchers do not face any difficulty for
obtaining literature in this field.
Key Words: Amplitude-Zone-Time-Epoch-Coding,
Coordinate-Reduction-Time-Encoding System , reconstructed signal.
Karhunen-Lodve transform, Percent root-mean-
square difference, Turning point .
1. Introduction ECG signals, normally
consist of 3 to 12 individual leads, each recording
11 bits per sample, 1000 samples per second, and
last 24 hours. When converted to a digital format,
this single ECG record requires a total of 1.36
gigabytes of computer storage. Considering the 10
million ECGs annually recorded for the purposes of
comparison and analysis. Furthermore, the
growing need for transmission of ECGs for remote
analysis is hindered by capacity of the average
analog telephone line and mobile radio. For this
purpose, some portable ECG sensing systems
associated with wireless data transmission have
been developed for real-time ECG signal recording.
As the sampling rate, sample resolution,
observation time and number of leads increase, the
amount of ECG data also increases and so the huge
storage capacity is required. Especially, in data
transmission, the amount of transmission time
increases and it needs more bandwidth for
compensation. With all of these limitations, ECG
data compression has been one of the most active
research areas in biomedical engineering and signal
processing. Evaluating an ECG data compression
algorithm requires some performance measures to
quantify the quality of the reconstructed ECG
signals. The most criteria used in the literature are
compression ratio (CR) and percent root-
meansquare difference (PRD). The CR is computed
from the ratio of original file size (in bits) to the
length of output bit stream (compressed signal)
Original samples in bytes
CR 
Compressed samples in bytes
The numerical distortion measure used is the
percent root-mean-square difference (PRD), which
is calculated by
PRD  i 1

  f i  fˆ i 
n
2
 f i 
n
2
i 1
Where f i  is the original signal and fˆ i  is the
2. Methodologies And Analysis
2.1. Time domain techniques These are direct
Signal compression techniques where the samples
of the signal are directly handled to provide the
compression. Coding by time-domain methods is
based on the idea of extracting a subset of
significant signal samples to represent the signal.
The reconstruction, in the decoder, is done by
performing linear interpolation between the
retained samples. In 1968, J. R. Cox [1], proposed a
Amplitude-Zone-Time-Epoch-Coding (AZTEC)
algorithm. AZTEC takes raw ECG data and
replaces sequences of sample points with plateaus
and slopes. For signals sampled at 200 sps with a
12-bit analog-to-digital converter, AZTEC's
average data-compression ratio is observed to be
about 80 percent (5 :1) and PRD of 31% . In 1978,
W. C. Mueller [ 2], suggested a new turning point
(TP) algorithm. This reduces the amount of data by
50 percent (2-to-1 reduction) and and PRD of
9.40%. In 1982, John P. Abenstein [3], presented
a new data-reduction algorithm called Coordinate-
Reduction- Time-Encoding System (CORTES).
This is a hybrid of the turning-point and AZTEC
algorithms. CORTES algorithm achieved data
reduction by 80 percent and PRD of 11%. As the
ECG is sampled, CORTES simultaneously applies
AZTEC to the isoelectric region of the ECG signal
and TP to the clinically significant higher
frequency regions of the ECG signal. In 1985, R.
C. Barr [4] presented FAN algorithm. In 1989,
George B. Moody[5] presented TRIM algorithm .In
1991, S. C. Tai [6] presented SLOPR algorithm. In
1992, Wilfried Philips [7], proposed method based
on high-degree polynomial expansions and got
better result than DCT. In 1993, Gil Nave [8],
proposed compression algorithm based on
subautoregression model (Long-Term Prediction
model). In 1993, S. C. Tai [9] again presented
AZTDIS algorithm for improving time domain
techniques.
Contrary to these above heuristic algorithms based
on time domain techniques,there are some
Optimization algorithms given below. In 1997, D.
Haugland [10 ],used optimisation and graph theory
to solve the ECG compression problem. In 1998,
Ranveig Nygaard [11] used optimization algorithm
such that reconstruction can be made by second
order polynomial interpolation in the extracted
signal samples. Time domain techniques can‘t
achieve higher compression ratios and are sensitive
to the noise contained in original ECG signals. This
problem is rectifying in Transformation
compression techniques.
2.2.Transformation compression techniques
These methods mainly utilize the spectral and
energy distributions of the signal. Generally this
means processing the input signal by means of
some transform, and properly encoding the
transformed output. Signal reconstruction is
achieved by an inverse transformation process. In
1986, B. R. Shankara Reddy [12],applied an
approach based on the method of Fourier
descriptors (FD's) technique for ECG data
compression. In 1990 C. P. Mammen
[13],presented a data-reduction algorithm
m-AZTEC, which is a multichannel extension of
the amplitude zone time epoch coding (AZTEC)
algorithm. He also used a scheme based on vector
quantization (VQ) to code correlated multichannel
signals. VQ is a coding technique, which has been
also employed to compress image and speech
signals [14]. Thus VQ-coded m-AZTEC achieved a
compression-factor of 2.0 over m-AZTEC, yielding
a net compression factor of 8.6. In 1993, A.Enis
Cetin[15], presented a new multilead ECG data
compression technique. In this technique sample of
thestandard ECG lead signals are first linearly
transformed. Then, resulting transform domain
signals are compressed using various coding
methods, including multirate signal processing and
transform coding methods. As a linear transformer,
both the optimum transform, Karhunen-Loeve,
and a suboptimum transform, DCT are used. In
1996, Salvador Olmos[16], proposed the ECG
compression strategy based on Karhunen-Lodve
transform technique. The KLT has been applied in
two different ways: firstly, using the entire beat as
pattern vector of the KLT, and the second one
using independent signal subwindows: P wave,
QRS complex and ST-T complex. The latter
method (KLW) achieves better compresion ratio of
17.2 and mean squared error (MSE) value of
0.3%. In 1995,Adrianus Djohan [17], proposed the
ECG compression strategy based on Discrete
symmetric wavelet transform [18],optimal bit
allocation and Huffman coding. Using this method
, he achieved better compression ratio of 8 to 1
with PRD =3.9 % for 117 in the MIT-BIH
Arrhythmia Database, in contrast to the AZTEC
compression ratio of 6.8 to 1 with PRD =10.0 %
and the fan algorithm compression ratio of 7.4 to 1
with PRD =8.7 % . In 1997, Michael L.
Hilton[19], proposed the ECG compression
strategy using Wavelet and wavelet packet-based
compression algorithms based on embedded zero
tree wavelet (EZW) coding. The algorithms are of
low computational complexity  n  and
 n log 2 n respectively. The wavelet packets of
Coifman and Meyer [20] are a generalization of the
wavelet transform that allow for arbitrary tree-
shaped bandpass filterings and can be adapted to
the characteristics of the particular signal being
analyzed . Hilton achieved better compression ratio
of 8:1 with PRD =2.6% for 117 in the MIT-BIH
Arrhythmia Database. The resulting PRD‘s by
wavelet packet were consistently higher (worse) for
the same signal when compared with those of the
plain wavelet coder. In 1997, Baohua Wang [21],
proposed Finite-State Vector Quantization and
directly employed VQ to compress a single-channel
ECG signal because the error the m-AZTEC
algorithm yields is relatively large and its anti-
noise capability is weak. In 1998,Jie Chen[22],
proposed the ECG compression strategy based on
orthonormal wavelet transform and an adaptive
quantization strategy, by which a predetermined
percent root mean square difference (PRD) can be
guaranteed with high compression ratio and low
implementation complexity. He achieved better
compression ratio of 8 .31 to 1 with PRD =1.07%.
In 2000, Shaou-Gang Miaou[23], proposed a new
adaptive vector quantization (AVQ) scheme to
exploit this correlation for multichanner ECG data
compression. The proposed approach has better
coding performance than using single channel
approach by 1.26 & 0.1 times.In 2000, E. Berti[
24], presents an electrocardiogram ECG data
reduction method based on Walsh spectrum double
logarithmic quantization. In 2000, Zhitao Lu[25] , bank with a parallel structure [31] that guarantees
proposed the ECG compression strategy based on the quality of the retrieved signal. This method
the set partitioning in hierarchical trees (SPIHT) utilises a nearly-perfect reconstruction cosine
compression algorithm. The SPIHT algorithm [26] modulated filter bank to split the incoming signals
had achieved notable success in still image coding into several subband signals that are then quantised
and modified to suit the special characteristics of through thresholding and Huffman encoded. He
ECG signals. achived compression ratio of 8.24 with percent root
mean square difference (PRD) of 1.176% for MIT-
―Table 1 . Average Test Results For Dataset‖ BIH 117 signal. In 2008, M. Sharafat Hossain[32],
proposed algorithm for an ECG compression
CR 4:1 5:1 8:1 10:1 12:1 16:1 20:1 system using wavelet transform and thresholding
PRD 1.19 1.56 2.46 2.96 3.57 4.85 6.49 technique based on energy packing efficiency
(EPE). As the WT decomposes the ECG signal into
―Table 2. PRD Comparison‖ multiresolution bands, a multi-level thresholding
strategy based on EPE is applied by him. The
Algorithm PD CR Signal Sampling Bits/sample algorithm can be tuned to required compression
R% R(HZ) ratio and PRD by selecting thresholds based on a
SPIHT 1.18 8:1 MIT- 360 11
BIH 117
desired EPE. He achived an average compression
Hilton 2.6 8:1 MIT- 360 11 ratio of 13.57 with percent root mean square
BIH 117 difference (PRD) of 4.87 for a large set of normal
Diohn 3.9 8:1 MIT- 360 11 and abnormal ECG signals extracted from MIT-
BIH 117
BIH database.
In 2001, Shaou-Gang Miaou[ 27], propose a MC ―Table 4. Result For MIT_BIH Arrhythmai Database‖
adaptive vector quantization (MC-AVQ) algorithm,
where AVQ is performed on each channel and Signal Threshold CR PRD %
across channels to exploit the correlation within a MIT-BIH 0.1364 15.12 2.33
channel and across channels. VQ often suffers from 117
MIT-BIH 0.2909 15.70 4.13
the mismatch problem, i.e., the codebook generated
119
from a set of training data is not suitable for the test
sets. Adaptive vector quantization (AVQ) is In 2009, N. Boukhennoufa[33], proposed method
designed to solve this problem by updated AVQ
combines the adapted SPIHT(Set Partitioning In
codebook according to changing characteristics of
Hierarchical Trees) method with VKTP (Vector K-
the test set. In 2002, Bashar A. Rajoub [28],used
Tree Partitioning) coder[34] and achived to
discrete wavelet transform (DWT) to compress
increase the performance of the compression
ECG signals. The ECG signal is first preprocessed methods and to reduce the reconstruction error
by normalization, mean removal, and zero padding. provided. He achieved better compression ratio of 8
Normalization and mean removal guarantees that
.11 to 1 with PRD =1.8% for 117 in the MIT-BIH
all significant coefficients be less than one. Thus,
Arrhythmia Database . In 2009, Gyu - Hyeok Jeong
reducing the number of bits needed to represent
[35],presented an improved wavelet compression
each coefficient. Zero padding reduces the
algorithm is proposed for electrocardiogram (ECG)
reconstruction errors at both ends of the signals which is combined with the lifting wavelet
compressed signal. Thresholding of the wavelet transform (WT) and the dynamic multi-stage vector
coefficients is done based on a desired energy
quantization (MS-VQ). In 2009, Iulian B.
packing efficiency (EPE).
Ciocoiu[36],proposed a novel compression
scheme based on combining DWT analysis with a
―Table 3. PRD Comparison‖
space-scale selective mask constructed following
Algorithm PDR CR Signal Sampling Bits/sa retina-like foveation principles. Foveation enables
% R(HZ) mple the definition of a proper mask that will modulate
DWT 1.06 22.19 MIT- 360 11 the coefficients given by the Discrete Wavelet
:1 BIH Transform of an ECG record. The mask is spatially
117
DWT 1.95 23:1 MIT- 360 11 selective and provides maximum accuracy around
BIH specific regions of interest. Subsequent denoising
119 and coefficient quantization is further combined
ASEC[29] 5.5 21.6: MIT- 360 11 with a lossless compression technique in order to
1 BIH
119
provide high compression ratios at low
reconstruction errors. He achieved better
In 2004, M. Blanco-Velasco[30 ],presemted an compression ratio of 5.55 to 1 with PRD = 0.52%
easy to use and efficient ECG compression scheme for 117 in the MIT-BIH Arrhythmia Database. In
based on an M-channel maximally decimated filter 2010, Ataollah Ebrahimzadeh[37], has performed
a different thresholding method using the Three- In 2008, Emna Zoghlami Ayari [49],applied
Level Quantization and used various wavelets and algorithm based on template model fitting by using
both EZW & Huffman encoding for a better a nonlinear least squares optimization procedure.
compression. He have evaluated the effect of signal His proposed a low computational cost model
length and iterations of quantization on the having only 12 parameters to fully represent the
compression ratio (CR). In low iterations, PRD & ECG signal. In 2008, Ibrahim Khalil[50],
CR increased and CC (correlation coefficient) decomposed ECG data first using wavelet packet
decreased, in high iterations of quantization values decomposition (WPD), and then low frequency
of PRD, CR decreased and CC increased. components can be given highest priority before
being transferred over wireless networks. His
―Table 5. The PRD, CR Comparison Of Different results show both high and medium priority ECG
Coding Algorithm For MIT-BIH 117‖ packets suffer no delays which is very beneficial
for ECG transmission system over bandwidth
ALGORITHMS PRD% CR constrained wireless networks. In 2006, Byung S.
Djohn[17] 3.9 8 Kim[51] , proposed a wavelet transform-based
Hilton[19] 2.6 8 ECG compression algorithm with low delay
Chen[22] 1.07 8.31 characteristics for instantaneous, continuous ECG
SPIHT[38] 1.18 8 transmission suitable for telecardiology
Benzid[39] 2.55 16.24 applications over wireless network. His proposed
JPEG2000[40] 1.03 10
algorithm reduces the frame size as much as
possible to minimize delay, whereas maintaining
WT&Huffman[41] 3.2 9.4
reconstructed signal quality without affecting the
Blanco-Velasco[30] 1.17 8.24 diagnostic characteristics of the ECG signals at the
Sec&SPIHT[42] 1.01 8 receiver end.
SPIHT&VQ[43] 1.45 8
N. Boukhennoufa[33] 1.18 8.11 4. Conclusion
A.Ebrahimzadeh[37] 0.97 13.92
ECG data compression has become in the last
3. Recent Work 40 years one of the most challenging topics in
Biomedical signal processing. However, there is
In 2010,Cheng-Tung Ku [44], proposed his 1-D still a continual need for the advancement of
RRO-NRDPWT, a new multivariable quantization algorithms adapted for ECG data compression.
algorithm to overcome the problem of Word- This paper is an effort for an overview of ECG data
length-growth (WLG) effect that arises in online compression , with a bibliographic survey of
applications. For efficient reconstruction quality relevant background, practical requirements, the
control, he Combining the reversible round-off present state, and techniques. The citations listed in
linear transformation (RROLT) theorem[45] and 1- this bibliography provide a representative sample
D nonrecursive discrete periodic wavelet transform of current thinking pertaining to ECG data
compression .The necessity of better ECG data
(NRDPWT) [45]. In 2009, Vibha
compression methods is even greater today than
Aggarwa[47],applier a new quality controlled just a few years ago for several reasons. The
Discrete Cosine transform (DCT) and Laplacian quantity of ECG records is increasing by the
Pyramid based compression method for millions each year, and previous records cannot be
electrocardiogram (ECG) signal. The transformed deleted since one of the most important uses of
coefficients are thresholded using the bisection ECG data is in the comparison of records obtained
algorithm in order to match the predefined user over a long range period of time. The ECG data
specified PRD within the tolerance. His results compression techniques are limited to the amount
show that Laplacian Pyramid provides better CR of time required for compression and
over 2.5 PRD value on the normal rhythm (MIT- reconstruction, the noise embedded in the raw ECG
BIH 117) as compared to DCT, whereas on signal, and the need for accurate reconstruction of
abnormal rhythm (MIT-BIH 232) the behavior of the P, Q, R, S, and T waves. With the current
the transforms is opposite. In 2009, Hyejung wireless technology the uplink bandwidth to
Kim[48], proposed quad level vector (QLV) for the transfer ECG data is limited and hence more
ECG holter system. He achived average algorithms for ECG Compression and its real-time
compression 16.9:1 with 0.641% PRD value and transmission is yet to be explore.
the encoding rate is 6.4 kbps. The accuracy
performance of the R-peak detection is 100%
without noise and 95.63% at the worst case with
−10-dB SNR noise. His proposed compression
techniques reduced the processing cost by 45.3%.
5. Reference
[1] J. R. Cox, F. M. Nolle, H. A. Fozzard, and G. C.
Oliver, T., "AZTEC: A preprocessing program for real-
time EGC rhythm analysis," IEEE Trans. Biomed. Eng.,
vol. BME-15, pp. 128-129, 1968
[2] W. C. Mueller, "Arrhythmia detection program for an
ambulatory ECG monitor," Biomed. Sci. Instrum., vol.
14, pp. 81-85, 1978.
[3] J. P. Aberstein and W. J. Tompkins," A new data
reduction algorithm for real time ECG analysis" IEEE
Trans Biomed . Eng. Vol. BME-29, pp. 43-48, 1982.
[4] D. A. Dipersio and R. C. Barr, ―Evaluation of the fan
method of adaptive sampling on human
electrocardiograms,‖ Med. Biol. Eng. Comput., pp. 401–
410, Sept. 1985
[5] G. B. Moody, R. G. Mark, and A. L. Goldberger,
―Evaluation of the TRIM ECG data compressor,‖ in
Proc. Comput. Cardiol., 1989, pp. 167–170.
[6] S. C. Tai, ―Slope—A real-time ECG data
compressor,‖ Med. Biol. Eng. Comput., vol. 29, pp. 175–
179, Mar. 1991.
[7] W. Philips and G. De Jonghe, ―Data compression of
ECG‘s by highdegree polynomial approximation,‖ IEEE
Trans. Biomed. Eng., vol. 39, no. 4, pp. 330–337, Apr.
1992.
[8] G. Nave and A. Cohen, ―ECG compression using
long-term prediction,‖ IEEE Trans. Biomed. Eng., vol.
40, pp. 877–885, Sept. 1993.
[9] S. C. Tai, ―AZTDIS—A two phase real-time ECG
data compressor,‖ J. Biomed. Eng., vol. 15, pp. 510–515,
Nov. 1993.
[10] D. Haugland, J. Heber, and J. Husøy, ―Optimization
algorithms for ECG data compressio,‖ Med. Biomed.
Eng. Computing, vol. 35, pp. 420–424, July 1997.
[11] R. Nygaard and D. Haugland, ―Compressing ECG
signals by piecewise polynomial approximation,‖ in
Proc. Int. Conf. Acoust. Speech, Signal Processing,
Seattle, Washington, May 1998, pp. 1809–1812..
[12] B.R.S. Reddy & I.S.N. Murthy, ―ECG Data
Compression Using Fourier Descriptors,‖IEEE Trans.
Biomed. Eng., vol. BME-33, pp. 428-433, April 1986.
[13] C. P. Mammen and B. Ramamurthi, ―Vector
quantization for compression of multichannel ECG,‖
IEEE Trans. Biomed. Eng., vol. 37, pp. 821–825, Sept.
1990.
[14] R. M. Gray, ―Vector quantization.‖ IEEEASSP
Mag.. vol. I , pp. 4-29, Apr. 1984.
[15] A.E. Cetin, H. Köymen & M.C. Aydin,
―Multichannel ECG Data Compression by Multirate
Signal Processing and Transform Domain Coding
Techniques,‖ IEEE Trans. Biomed. Eng., vol. 40,
pp.495-499, May 1993.
[16] S. Olmos, M. Millan, J. Garcia & P. Laguna, ―ECG
Data Compression with the Karhunen Loève Transform,‖
Comput. Cardiol. IEEE, pp. 253-256, Sept.1996.
[17] Djohan, A., Nguyen, T.Q., Tompkins, w. J., ECG
compression using discrete symmetric wavelet transform,
IEEE 171h annual Conference, Engineering in Medicine
and Biology Society, voU, pp. 167-168, 1997.
[18] I . Daubechies, ―Ten Lectures on Wavelet,‖ CBMS
Conference SIAM , Philadelphia, 1992.
[19] M. L. Hilton, ―Wavelet and Wavelet Packet
Compression of Electrocardiograms,‖ IEEE Trans.
Biomed. Eng., vol. 44, pp. 394-402, 1997.
[20] R. R. Coifman and Y. Meyer, ―Orthonormal wave
packet bases,‖ Math. Dept., Yale Univ., New Haven, CT,
Tech. Rep., 1990.
[ 21] W. Baohua and Y. Guoxin, ―Compression of ECG
data by vector quantization,‖ IEEE Eng. Med. Biology
Mag., vol. 16, no. 4, pp. 23–26, 1997.
[22] J. Chen, and S. Itoh, ―A wavelet transform-based
ECG compression method guaranteeing desired signal
quality,‖ IEEE Trans. Biomed. Eng., vol. 45, pp. 1414–
1419, Dec. 1998.
[23] Shaou-Gang Miaou', Heng-Lin Yen', and Chia-Yang
Lin', ―Multi-Channel ECG Data Compression Using
Two-Stage Adaptive Vector Quantization,‖ Annual
EMBS International Conference, July 23-28, 2000,
Chicago IL
[24] E. Berti & f. Chiaraluce, ―Reduction of Walsh
Transformed Electrocardiograms by Double Logarithmic
Coding,‖ IEEE Trans Biomed. Eng., vol. 47, pp. 1543-
1547, Nov. 2000.
[25] Z. Lu, D. Y. Kim, and W. A Pearlman, "Wavelet
Compression of ECG signal by the Set Partitioning in
Hierarchical Trees (SPIRT) algorithm, " IEEE
Transactions on Biomedical Engineering, Vol. 47,
July 2000 , pp. 849-856.
[26] A. Said and W. A. Pearlman, ―A new, fast and
efficient image codec based on set partitioning in
hierarchical trees,‖ IEEE Trans. Circuits Syst. II, vol. 6,
pp. 243–250, June 1996.
[27] Shaou-Gang Miaou and Heng-Lin Yen,
―Multichannel ECG Compression Using Multichannel
Adaptive Vector Quantization,‖ @ 2001 IEEE.
[28] B. A. Rajoub, ―An efficient coding algorithm for the
compression of ECG signals using the wavelet
transform,‖ IEEE Trans. Biomed. Eng., vol. 49, no. 4, pp.
355–362, Apr. 2002
[29] S. Jalaleddine, C. Hutchens, R. Strattan, and W.
Coberly, ―ECG data compression techniques—A unified
approach,‖ IEEE Trans. Biomed. Eng., vol. 37, pp. 329–
343, Apr. 1990.
[30] M. Blanco-Velasco, F. Cruz-Roldan, J. I. Godino-
Llorente, and K. E. Barner, ―ECGcompression with
retrieved quality guaranteed,‖ Electron. Lett., vol. 40, no.
23, pp. 1466–1467, Nov. 11, 2004.
[31] Strang, G., and Nguyen, T.Q.: ‗Wavelets and filter
banks‘ (Wellesley, Cambridge, MA, 1996).
[32] M. Sharafat Hossain, Tareq Aziz, M.A. Haque,
―ECG Compression Using Multilevel Thresholding of
Wavelet Coefficients,‖@ 2008 IEEE .
[33] N. Boukhennoufa 1, K. Benmahammed 2, M. A.
Abdi 1 and F. Djeffal ―Wavelet-Based ECG signals
compression using SPIHT technique and VKTP coder ―
978-1-4244-4398-7/09/$25.00 ©2009 IEEE
[34] J. Martinez-Alajarin, J. Martinez-Rosso, and R.
Ruiz-Merino, ―Encoding technique for binary sequences
using vector tree partitioning applied to compression of
phonocardiographic signals,‖ Electron. Lett., vol. 44, pp.
84-85, 2008.
[ 35] Gyu-Hyeok Jeong & In-Sung Lee, ―Wavelet-based
ECG Compression using Dynamic Multi-stage Vector
Quantization,‖@ ICIEA 2009
[36] Iulian B. Ciocoiu, ―ECG Signal Compression Using
Wavelet Foveation,‖ @2009 IEEEX
[37]Ataollah Ebrahimzadeh , Milad Azarbad ―ECG
Compression Using Wavelet Transform and Three-Level
Quantization ― @2010 IEEE
[38] Z. Lu, D. Y. Kim, and W. A Pearlman, "Wavelet
Compression of ECG signal by the Set Partitioning in
Hierarchical Trees (SPIRT) algorithm, " IEEE
Transactions on Biomedical Engineering, Vol. 47, July
2000 , pp. 849-856.
[39]R. Benzid, F. Marir, A. Boussaad, M. Benyoucef,
and D. Arar, ―Fixed percentage of wavelet coefficients to
be zeroed for ECG compression,‖ Electron. Lett., vol. 39,
pp. 830-831, 2003.
[40]Ali Bilgin, Michael W. Marcellin, Maria I. Altbach, "
Compression of Electrocardiogram Signals Using
JPEG2000 " ,IEEE Transactions on Consumer
Electronics, Vol. 49, No. 4, November, 2003, pp. 833-
840.
[41] A. Al-shrouf, M. Abo-Zahhad, S. M. Ahmed. "A
novel compression algorithm for electrocardiogram
signal based on the linear predictionof the wavelet
coefficients ", Digital Signal Processing, VoU3, 2003,
pp. 604-622.
[42] I. M. Rezazadeh. M. H.Moradi, AM. Nasrabadi, "
Implementing of SPIRT and subband Energy
Compression (SEC) Method on Two Dimensional ECG
Compression: A Novel Approach. Engineering in
medicine and biology, 27"'annual conference, IEEE,
China, 2005.
[43] S. M. E Sahraeian and E. Fatemizadeh. "Wavelet-
Based 2-D ECG Data Compression Method Using
SPIHT and VQ Coding ". IEEE Int. Conf on " Computer
as a tool" . EUOROCON 2007.
[44]Cheng-Tung Ku, King-Chu Hung , Tsung-Ching
Wu, Huan-Sheng Wang , ―Wavelet-Based ECG Data
Compression System With Linear Quality Control
Scheme,‖ @ 2010 IEEE
[45] C.-T. Ku, ―A high efficient ECG data compression
system based on reversible round-off discretewavelet
transform ,‖ Doctoral Dissertation, Inst. Eng. Sci.
Technol., Nat. Kaohsiung First Univ. Sci. Technol.,
Taiwan,May 2007.
[46] C. F. Tsai, H. S. Wang, and K. C. Hung, ―A high
efficient nonrecursive discrete periodized wavelet
transform for extracting the transformed coefficients of
coarser resolution levels,‖ in Proc. 2004 IEEE Asia-
Pacific Conf. Circuits Syst., Dec. 6–9, pp. 661–664.
[47] Vibha Aggarwal and Manjeet Singh Patterh,
―Quality Controlled ECG Compression using Discrete
Cosine Transform (DCT) and Laplacian Pyramid (LP),‖
@2009 IEEE
[ 48] Hyejung Kim , Refet Firat Yazicioglu, Patrick
Merken, Chris Van Hoof and Hoi-Jun Yoo, ―ECG Signal
Compression and Classification Algorithm With Quad
Level Vector for ECG Holter System,‖ @2009 IEEE
[49] Emna Zoghlami Ayari, Rheinhardt Tielert and
Norbert Wehn, ―Template-based Compression of ECG
Signals,‖@ 2008 IEEE
[50] Ibrahim Khalil abd Fahim Sufi, ―Real-Time ECG
Data Transmission with Wavelet Packet Decomposition
over Wireless Network,‖ @ 2008 IEEE.
[51] Byung S. Kim, Sun K. Yoo, and Moon H. Lee ,
―Wavelet-Based Low-Delay ECG Compression
Algorithm for Continuous ECG Transmission‖ , IEEE
Transactions on Information Technology in Biomedical
Vol . 10, No. 1, JANUARY 2006