Professional Documents
Culture Documents
Empirical Determination of an
ECG Compression Ratio for Mobile
Telecardiology Applications
Sun Kook Yoo, Ph.D.,2,4,5,6,7 Kwanghyun Lee,1 and Moon H. Lee, M.D.3 for example, allowing the monitoring of emergency patients during
ambulance dispatch.1 Despite advances in network technology, it is
1
Yonsei University College of Medicine, Seoul, South Korea. not always possible to secure sufficient network bandwidth for real-
2
Medical Engineering, 3Internal Medicine, Yonsei University, time continuous ECG transmission,2 and bandwidth is even more
College of Medicine, Seoul, South Korea. limited in a wireless network.
4
Brain Korea 21 Projects for Medical Science, 5Center for Emergency In order to work within limited bandwidth, it becomes necessary
Medical Informatics, 6Human Identification Research Center, 7Center to compress the ECG signal. Many ECG compression techniques have
for Signal Processing Research, Yonsei University, Seoul, South Korea. been studied to maximize compression performance and comply
with bandwidth limitation,3,4 but few studies have focused on the
Abstract permissible amount of ECG compression associated with mobile
It is important to consider electrocardiogram (ECG) data compression communication. Compression performance is generally measured
that does not sacrifice diagnostic quality significantly before apply- using an objective index, which calculates the squared difference
ing ECG data compression to mobile telecardiology applications. In between the original and the reconstructed ECG signals.5 However,
this paper, we assessed the reconstructed ECG quality after compres- that objective index using rate-distortion criteria alone cannot reveal
sion with a wavelet-based low-delay algorithm, using both subjective the diagnostic relevance of the compression ratio (CR). Alternatively,
and objective indices. We included diverse ECG databases including cardiologists can evaluate compression performance subjectively by
both normal and abnormal ECG data, and evaluated the relationship visually ranking the reconstructed ECG quality. That visual ranking
between the subjective and objective indices, paying close attention to can be used as a subjective index describing the clinical adequacy of
specific cases in which there was a large discrepancy between the objec- the reconstructed ECG quality from the cardiologist’s point of view.6
tive and subjective quality. Based on our observations, an empirically However, visual inspection is constrained by the limited availability
determined compression ratio can be applied to compress continuous of cardiologists.
ECG signals in limited-bandwidth mobile telecardiology applications. The use of both objective and subjective indices can overcome
the limitations inherent in each index alone. Thus, a permissible CR
Key words: electrocardiogram, data compression, telecardiology, that does not sacrifice diagnostic performance significantly can be
compression ratio determined before applying the ECG compression technique to a tele-
cardiology application. In this paper, ECG compression performance
Introduction is analyzed in terms of CR by using the wavelet-based low-delay ECG
T
he electrocardiogram (ECG) is essential in diagnosing patients compression algorithm,2 which was chosen as a typical technique
with heart problems. In remote or underserved areas without suitable for continuous ECG transmission over a mobile network.
access to cardiologists, an ECG can be monitored through Measurements were obtained using various ECG databases includ-
a telecardiology system, which transmits the ECG signal ing both normal and abnormal ECG data. In addition, extraordinary
from the patient to a remote site through a network. In addition, cases, in which the objective index does not match the subjective
telecardiology may also be used to achieve continuous, instanta- index, were comprehensively inspected to determine the permissible
neous transmission of an ECG signal in real-time during movement, CR before transmitting the compressed ECG over a mobile network.
(a) Uncompressed TYP (b) Compressed TYP (c) Uncompressed SVA (d) Compressed SVA
FIG. 1. Compressed electrocardiogram (ECG) signals with lowest changing ratio of objective performance (CROP) typical testing data set (TYP)
and highest CROP supraventricular arrhythmia (SVA) are compared with uncompressed ECG signals.
© M ARY ANN LI E BE RT, I NC. • VOL. 14 NO. 2 • M ARCH 2008 TELEMEDICINE and e-HEALTH 157
YOO ET AL.
CR vs. RMSE
60
50
40
RMSE
30
20
10
0
0 5 10 15 20 25 30
CR
FIG. 2. The relationship between root mean squared error (RMSE) and compression ratio (CR) for all testing data.
original ECG and the reconstructed ECG. For N ECG samples, the reconstructed ECG, respectively. The subjective index used the Mean
RMSE is defined as Opinion Score (MOS) to quantify cardiologist assessment of reconstructed
ECG quality.6 The MOS provides a numerical interpretation of the per-
RMSE = N1 Σf(i) ⫺ fˆ(i)
Ν
2
ceived quality of the reconstructed ECG after compression. In the MOS
measurement, 3 cardiologists were blinded and asked to score the clinical
where f(i) and fˆ(i) are the ith samples of the original ECG and the acceptability of a reconstructed ECG signal in comparison to the original
CR vs. MOS
TYP
6
VTA
5
ARR
4
MOS
MVE
3
NOR
2
SVA
1
MOR
0
SCD
0 5 10 15 20 25 30
CR
FIG. 3. The relationship between mean opinion score (MOS) and compression ratio (CR) for all testing data.
3
2
1
0
0 10 20 30 40 50 60
RMSE
FIG. 4. The relationship between root mean squared error (RMSE) and mean opinion score (MOS).
© M ARY ANN LI E BE RT, I NC. • VOL. 14 NO. 2 • M ARCH 2008 TELEMEDICINE and e-HEALTH 159
YOO ET AL.
SUBJECTIVE MEASUREMENT tive and objective qualities. Therefore, the permissible CR cannot be
As shown in Figure 3, for all testing data except the TYP set, the determined from either the MOS or RMSE criteria alone. Instead, the
MOS decreases as the CR increases. Generally, an increase in CR lower CR (10:1), between the subjective permissible CR (13:1) and the
causes deterioration in the visual quality of the reconstructed ECG. objective permissible CR (10:1), should be selected as the permissible
However, for the TYP data set, an increase in CR less than 30 has little CR for mobile telecardiology.
effect on the reconstructed visual quality, because the TYP data set
represents typical testing ECG with little variation. The TYP data set is CASES WITH GOOD OBJECTIVE QUALITY BUT BAD
relatively insensitive to changes in CR. When an MOS of 4 (a nearly SUBJECTIVE QUALITY
unnoticeable difference in ECG quality) is chosen as the subjective From a clinical point of view, the maintenance of subjective
threshold, the permissible subjective CR is found to be 13:1, which is quality (diagnostic quality) is more stringent than that of objective
the lowest among CRs for all testing data. The CRs corresponding to quality.5 Some ECG data show good objective quality (low RMSE),
an MOS of 4 are 13:1, 13:1, 19:1, 23:1, 26:1, 281:1, and 29:1, for the but bad subjective quality (low MOS). Even though the mean differ-
NOR, SVA, MOR, SCD, MVE, VTA, and ARR data sets, respectively. ence between the original ECG and the reconstructed ECG is small,
clinically important data along the time-sequenced ECG trace can be
RELATIONSHIP BETWEEN SUBJECTIVE AND OBJECTIVE significantly distorted, which can result in a low MOS despite the low
MEASUREMENTS RMSE. In order to clarify the relationship between the RMSE and the
As shown in Figure 4, the MOS (subjective index) correlates with MOS, the cases with an RMSE less than 10 and an MOS of 2 were
the RMSE (objective index) with the relationship MOS = ⫺0.0999 * examined.
RMSE ⫹5.168. However, there is large variation (0.525), even though Case 1 (Fig. 5): ARR 104: CR of 26:1, RMSE of 8.84, and MOS
the t-test (reliability of 95%) reveals a correlation (r = 0.725) between of 2. The reconstructed signal shows a shorter PR interval and some
the RMSE and MOS. This large variation between the subjective and P waves that were absent on the original. Thus, patients with heart
objective indices indicates that there is a discrepancy between subjec- problems may be misdiagnosed as normal.
Case 2 (Fig. 6): MVE 602: CR of 26:1, RMSE of 9.85, and MOS signal would still indicate a conduction disorder, but the cause of the
of 2. Although there are peaking P waves on the original, the recon- problem may be misdiagnosed. Altered P-wave appearance is also
structed P waves appear normal. A peaking P wave is one of the found.
diagnostic indicators of an abnormal right heart. Therefore, because Case 7 (Fig. 11): MOR 15814 for CR of 26:1, RMSE of 10.00, and
of the altered P-wave appearance, patients with a right cardiac prob- MOS of 2. The energy of the P wave is lower and the appearance has
lem may be missed. also widened on the reconstructed image. The reconstructed PR inter-
Case 3 (Fig. 7): NOR for CR of 28:1, RMSE of 7.22, and MOS of 2. val is shorter than the original. These could lead to the misdiagnosis
PR intervals are longer than in the original. Additionally, the T waves of a healthy individual.
appear variable in the reconstructed signal, although all the T waves
in the original share the same features. Discussion
Case 4 (Fig. 8): MOR 114 for CR of 26:1, RMSE of 9.77, and MOS In order to cope with bandwidth limitations, compression of the
of 2. There are some shortened PR intervals on the reconstructed ECG signal is inevitable. Because lossy compression can achieve
signal. The energy of the P wave is smaller and the intervals are lon- a higher CR than can lossless compression, the lossy compression
ger than in the original. Also, there is an initial blurring in the QRS is more suitable for low-bandwidth telecardiology despite the
complex. These changes may lead to the misdiagnosis of a healthy distortion it produces.11 In general, the greater the CR applied to
person as abnormal. the original ECG, the lower the quality of the reconstructed image.
Case 5 (Fig. 9): MOR 116 for CR of 19:1, RMSE of 6.04, and MOS Because of this tradeoff, the CR should be carefully selected when
of 2. The P wave on the original has disappeared and an additional P allocating the ECG transmission bitrates for a given network band-
wave has been generated in the PR interval. These factors may lead width. When allocating bits, the most commonly used measure is
to incorrect diagnosis in patients with heart problems. the RMSE (objective measure). However, this measure does not
Case 6 (Fig. 10): MOR 232 for CR of 19:1, RMSE of 8.88, and reflect the diagnostic value of the compressed ECG. The diagnostic
MOS of 2. A newly created abnormal P wave on the reconstructed quality can be validated precisely by expert cardiologists (subjec-
© M ARY ANN LI E BE RT, I NC. • VOL. 14 NO. 2 • M ARCH 2008 TELEMEDICINE and e-HEALTH 161
YOO ET AL.
tive measure). However, cardiologists are not always available. databases used should contain diverse P waves and ST segments,
As discussed in the Results section, the RMSE mostly correlates since P wave recognition (generally the smallest wave in the ECG
with diagnostic quality. However, the subjective measure cannot signal) is a difficult task in the diagnosis of ECG abnormalities,
be consistently substituted by the objective measure. Each portion and the ST segment is an important feature identifying myocardial
of the ECG waveform, such as the P-wave, QRS-complex, ST seg- infarction.12,13
ment, T-wave, and QT-interval, has its own diagnostic meaning. The subjective and objective quality of the reconstructed ECG
RMSE, as a mean performance measure, considers the error in all can be affected by the compression technique employed.3,4 In this
portions of the ECG waveform equally, and does not especially study, a wavelet-transform low-delay ECG compression algorithm
reflect the quality of those portions that are most diagnostically was selected for a telecardiology application over a mobile network
salient. Nonstationary ECG waveforms, missing beats, and noise- with limited bandwidth, which required a continuous ECG transmis-
to-signal problems can also affect the reconstructed ECG quality, sion with little delay. These same procedures should be applicable to
which can lead to a large variation in the correlation between different ECG compression algorithms. Nevertheless, further research
objective and subjective measures. is required, including more patterns of ECG data and more cardiolo-
The permissible CR should be determined as the lowest value gist involvement to reduce bias and to generalize quality evaluation
that satisfies both the objective and subjective criteria. The great- procedures.
est obstacle in this determination is the variability of ECG data, In this study, we used both subjective and objective indices
both in the case of the individual patient over a period of time to determine the permissible CR for telecardiology applications
and with respect to different patients. Accordingly, both criteria over a mobile network that required continuous ECG transmission
cannot be adequately evaluated using only specific patient data with little delay. The combination of both indices can overcome
with a limited observation interval. Instead, in our study, the CR the limitations inherent in each index alone. By using diverse
was empirically determined using diverse ECG databases, includ- patterns of ECG testing data and a wavelet-transform-based low-
ing both normal and abnormal ECG data. Importantly, the ECG delay ECG compression algorithm, the reconstructed ECG qual-
FIG. 9. Morphological testing data set 116 for compression ratio of FIG. 10. Morphological testing data set 232 for compression ratio of
19:1 (a) original and (b) reconstructed electroencephalograms (ECGs). 19:1 (a) original and (b) reconstructed electroencephalograms (ECGs).
REFERENCES
1. Alesanco A, Olmos S, Istepanian RS, Garcia J. Enhanced real-time ECG coder for pack-
etized telecardiology applications. IEEE Trans Inf Technol Biomed 2006;10:229–236.
2. Kim BS, Yoo SK, Lee MH. Wavelet-based low-delay ECG compression algorithm
for continuous ECG transmission. IEEE Trans Inf Technol Biomed 2006;10:77–83.
3. Rajoub BA. An efficient coding algorithm for the compression of ECG signals
using the wavelet transform. IEEE Trans Biomed Eng 2002;49:355–362.
4. Hilton ML. Wavelet and wavelet packet compression of electrocardiograms. IEEE
Trans Biomed Eng 1997;44:394–402.
5. Al-Fahoum AS. Quality assessment of ECG compression techniques using a wave-
let-based diagnostic measure. IEEE Trans Inf Technol Biomed 2006;10:182–191.
6. Al-Fahoum AS, Reza AM. Combined edge crispiness and statistical differ-
(a) Original encing for deblocking JPEG compressed images. IEEE Trans Image Process
2001;10:1288–1298.
7. Goldberger AL, Amaral LA, Glass L, Hausdorff JM, Ivanov PC, Mark RG, Mietus
JE, Moody GB, Peng CK, Stanley HE. PhysioBank, PhysioToolkit, and Physionet:
Components of a new research resource for complex physiologic signals.
Circulation 2000;101:215–220.
8. Taddei A, Distante G, Emdin M, Pisani P, Moody GB, Zeelenberg C, Marchesi C. The
European ST-T Database: Standard for evaluating systems for the analysis of ST-T
changes in ambulatory electrocardiography. Eur Heart J 1992;13:1164–1172.
9. Laguna P, Mark RG, Goldberger A, Moody GB. A database for evaluation of the
algorithms for measurement of the QT and other waveform intervals in the ECG.
Comput Cardiol 1997;24:673–676.
10. PhysioNet: Research Resource for Complex Physiologic Signals. Available at:
http://www.physionet.org (Last accessed 2005).
(b) Reconstructed ECG 11. Abo-Zahhad M, Rajoub BA. An effective coding technique for the compres-
sion of one-dimensional signals using wavelet transforms. Med Eng Phys
2002;24:185–199.
FIG. 11. Morphological testing data set 15814 for compression ratio of
26:1 (a) original and (b) reconstructed electroencephalograms (ECGs). 12. Khadra L, Al-Fahoum AS, Al-Nashash H. Detection of life threatening car-
diac arrhythmias using the wavelet transformation. Med Biol Eng Comput
1997;35:626–632.
13. Al-Fahoum AS, Howitt I. Combined wavelet transformation and radial basis
ity was inspected by clinicians, allowing a comparison between neural networks for classifying life-threatening cardiac arrhythmias. Med Biol
subjective and objective measures of quality. We observed that Eng Comput 1999;37:566–573.
the reconstructed ECG quality generally correlated inversely with
the CR, with some variation. However, some extraordinary cases Address reprint requests to:
demonstrated good objective quality but bad subjective quality at Sun K. Yoo
a high CR. Based on our observations, the empirically determined Medical Engineering
CR of 10:1 is the maximum compression ratio that should be used Yonsei University College of Medicine
to compress continuous ECG signals in limited bandwidth applica- Sudaemoon-gu Shinchon-dong
tions with little delay. Seoul 120-752, South Korea
© M ARY ANN LI E BE RT, I NC. • VOL. 14 NO. 2 • M ARCH 2008 TELEMEDICINE and e-HEALTH 163