Review
Wavelet-based electrocardiogram signal compression methods and their performances: A prospective review

https://doi.org/10.1016/j.bspc.2014.07.002Get rights and content

Abstract

Cardiovascular disease (CVD) is one of the most widespread health problems with unpredictable and life-threatening consequences. The electrocardiogram (ECG) is commonly recorded for computer-aided CVD diagnosis, human emotion recognition and person authentication systems. For effective detection and diagnosis of cardiac diseases, the ECG signals are continuously recorded, processed, stored, and transmitted via wire/wireless communication networks. But long-term continuous cardiac monitoring system generates huge volume of ECG data daily. Therefore, a reliable and efficient ECG signal compression method is highly demanded to meet the real-time constraints including limited channel capacity, memory and battery-power of remote cardiac monitoring, ECG record management and telecardiology systems. In such scenarios, the main objective of the ECG signal compression is to reduce the data rate for effective transmission and/or storage purposes without significantly distorting the clinical features of different kinds of PQRST morphologies contained in the recorded ECG signal. Numerous ECG compression methods have been proposed by exploiting the intra-beat correlation, inter-beat correlation and intra-channel correlation of the ECG signals. This paper presents a prospective review of wavelet-based ECG compression methods and their performances based upon findings obtained from various experiments conducted using both clean and noisy ECG signals. This paper briefly describes different kinds of compression techniques used in the one-dimensional wavelet-based ECG compression methods. Then, the performance of each of the wavelet-based compression methods is tested and validated using the standard MIT-BIH arrhythmia databases and performance metrics. The pros and cons of different wavelet-based compression methods are demonstrated based upon the experimental results. Finally, various practical issues involved in the validation procedures, reconstructed signal quality assessment, and performance comparisons are highlighted by considering the future research studies based on the recent powerful digital signal processing techniques and computing platform.

Introduction

Recent developments in digital technologies, sensors, efficient signal processing tools and wireless communication technologies enables to store and transmit biomedical signals for diagnosing patients diseases. Nowadays cardiovascular disease (CVD) is one of the most widespread health problems with unpredictable and life-threatening consequences in developing and developed countries in the world. The electrocardiogram (ECG) signals are widely used for detecting different kinds of heart diseases [1], [2], [3], [4], [5], [6], [7], [8]. For effectively detecting and diagnosing CVD problems, long-term multi-channel ECG signals are recorded with different sampling rates and bit resolutions. The ECG signals, for example, received from recording systems such as 12-lead ECG, the vector cardiography (VCG) high-resolution ECG, exercise ECG are digitized at the sampling rate ranging from 100 to 1000 Hz with resolution between 8 and 12 bit per sample [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18]. The amplitude of the recorded ECG signal on the skin is from 0.1 mV to 5 mV [1], [2], [3], [4], [5], [6], [7], [8], [9], [10]. The recording of the electrical field generated by the His and Purkinje activities produces a signal in the ECG with an amplitude range of about 1–10 μV which is useful for identification of conduction abnormalities [2]. The frequency extends from 0.05 Hz to 130 Hz [3], [10], [11], [12]. In ECG signal, notches and slurs may be superimposed on the slowly varying QRS complexes. The recommendations of the committee on electrocardiography of the American heart association suggest a conversion rate of 500 Hz with a 9-bit resolution [1]. The amount of ECG data depends upon the sampling frequency, sample width, the number of leads and record duration. Thus, the long-term ECG monitoring system generates large amount of cardiac data. Recently, miniaturized wearable ECG recorder enables both recording and transmission of ECG data via well-established telecommunication networks to the specialist diagnostic center [13], [14], [15]. In practice, sample rates from 100 Hz to 1000 Hz are used with 8–16 bit resolution per sample [1], [10], [16], [17], [18]. The data rate has become approximately 11–22 Mbits/h/lead. For the CVD analysis, multi-channel ECG signals are continuously recorded for 12–72 h to monitor ischemia, ventricular and supra-ventricular dysrhythmias, conduction abnormalities, QT interval and heart rate variability [2], [3], [4], [5], [6], [7], [8]. Hence, ECG record management system and telecardiology application requires reliable and efficient compression techniques for storage and continuous transmission purposes [34].

In general, single-channel and multi-channel ECG signal recording systems are used for effective analysis and diagnosis of cardiovascular diseases. Under long-term multi-channel recording conditions, ECG signals may exhibit three types of signal correlations: the intra-beat, the inter-beat and the inter-channel/lead. Fig. 1, Fig. 2, Fig. 3 illustrate different types of signal correlations. The intra-beat correlation represents the correlation between the successive samples in an ECG cycle. The inter-beat correlation represents the correlation between successive beats in a single-channel ECG signal. The correlation that exists between the signals from different channels is termed as inter-channel correlation. For effective storage and transmission purposes, there is a need for ECG data compression without introducing clinical distortion. The data compression is a method which is mainly developed: to reduce the channel bandwidth required for sending a given amount of information in a given time; to reduce the time required for transmitting a given amount of information in a given channel capacity [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37]. In general, the long-term multi-channel ECG signals have three correlations including inter-sample correlation within ECG beat, inter-beat correlation and inter-channel correlation. The ECG signals may contain long or short isoelectric regions between two local events or waves. The signal redundancy can be exploited when successive ECG samples are statistically dependent and the quantized ECG sample amplitudes occur with unequal probability [19], [29], [34]. In most ECG compression methods, the signal irrelevancy and redundancy properties in the time-domain, frequency-domain and time-frequency domain are exploited by using different signal processing techniques.

Section snippets

Compression efficiency

In most methods, the compression efficiency is evaluated in terms of the following performance measures: the sample reduction ratio (SRR), the compression ratio (CR), the compressed data rate (CDR) and the decoding rate (DR) [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35]. The sample reduction ratio (SRR) is defined as

SRR=number of samples within the input block (No)number of retained samples within the processed block (Nc)The compression measure

Classification of ECG compression methods

A digital information source is said to possess redundancy if either [20]: the symbols are not equally likely, or the symbols are not statistically independent. All data compression methods seek to minimize data storage by reducing the redundancy [21]. Redundancy in a set of discrete samples exists when signal samples are statistically dependent [27]. Data compression methods, which are successfully used for speech, image and video signals, can be employed for ECG signals. ECG data compression

Wavelet transform

In many signal processing applications, mathematical transformations are applied to a signal in order to extract information which may not be readily available from the original signal. A wavelet based signal processing technique is an effective tool for nonstationary ECG signal analysis and characterization of local waves (P, T and QRS complex morphologies). The details of design of wavelets and wavelet transforms can be found in [176], [177], [178], [179], [180], [181], [182], [183]. In this

One-dimensional wavelet based ECG compression methods

In ECG data compression, the wavelet and the wavelet packet transform are used to exploit the redundancy in the signal. After the signal is transformed into the wavelet domain, many coefficients are so small that no significant information is lost in the signal by setting these coefficients to zero [142]. Many 1-D and 2-D wavelet based compression methods are reported in literature. The 2-D wavelet-based ECG compression method achieved high compression ratio with low reconstruction error at the

Performance evaluation of ECG compression methods

In this section, the compression issues in the reported methods are described and investigated with different sets of experiments. In [136], [138], [172], various tests were carried out using the mita database and the PRD measurement criteria to evaluate the quality of the compressed signal and to select optimal coding parameters. In the next subsections, experimental results are presented for different signal processing techniques used in the ECG compression methods.

Discussion and concluding remarks

From the literature survey, many ECG compression methods have been proposed by exploiting the short-term correlation between adjacent samples, the considerable similarity between adjacent ECG beats (long-term correlation) and between the intra-lead ECG beats. In recent years, the wavelet transform has proven to be an effective tool in noise reduction, clinical features extraction and particularly ECG data compression because of its interesting properties such as time-frequency localization,

References (204)

  • H.A.M. Al-Nashash

    A dynamic fourier series for the compression of ECG using FFT and adaptive coefficient estimation

    Med. Eng. Phys.

    (1995)
  • U.C. Niranjan et al.

    ECG component delineation by Prony's method

    Signal Process.

    (1993)
  • B. Madhukar et al.

    ECG data compression by modeling

    Comput. Biomed. Res.

    (1993)
  • L.V. Batista et al.

    Compression of ECG signals by optimized quantization of discrete cosine transform coefficients

    Med. Eng. Phys.

    (2001)
  • R. Benzid et al.

    Constrained ECG compression algorithm using the block-based discrete cosine transform

    Digit. Signal Process.

    (2008)
  • Report of committee on electrocardiography, American Heart Association

    Recommendations for standardization of leads and of specifications for instruments in electrocardiography and vectorcardiography

    Circulation

    (1967)
  • F. Morris et al.

    ABC of Clinical Electrocardiography

    (2003)
  • W.J. Tompkins

    Biomedical Digital Signal Processing

    (1993)
  • M. Rangayyan Rangaraj

    Biomedical Signal Analysis: A Case-study Approach

    (2001)
  • R. McFee et al.

    Research in electrocardiography and magnetocardiography

    Proc. IEEE

    (1972)
  • D.B. Geselowitz

    Origin of the electrocardiogram

    IEEE Eng. Med. Biol. Mag.

    (1994)
  • A. Mattu et al.

    ECGs for the Emergency Physician

    (2003)
  • H.E. Awtry et al.

    Blueprints in Cardiology

    (2003)
  • M. Sabarimalai Manikandan

    Cardiovascular Signal Compression Using a New Wavelet Energy based Diagnostic Distortion Measure

    (2009, September)
  • László Gerencsér et al.

    The use of the SPSA method in ECG analysis

    IEEE Trans. Biomed. Eng.

    (2002)
  • N.V. Thakor et al.

    Estimation of QRS complex power spectra for design of a QRS filter

    IEEE Trans. Biomed. Eng.

    (1984)
  • G.J. Guang et al.

    High-frequency electrocardiogram analyzer

    IEEE Trans. Biomed. Eng.

    (1986)
  • M. Sabarimalai Manikandan et al.

    Wavelet-based ECG and PCG signals compression technique for mobile telemedicine

  • M. Sabarimalai Manikandan et al.

    Wavelet-threshold based ECG compression with smooth retrieved quality for telecardiology

  • B. Baby et al.

    Automated cardiac event change detection for continuous remote patient monitoring Devices

  • G.B. Moody et al.

    The MIT-BIH arrhythmia database in CD-ROM and software for use with it

  • G.B. Moody et al.

    The impact of the MIT-BIH arrhythmia database

    IEEE Eng. Med. Biol. Mag.

    (2001)
  • G.B. Moody et al.

    PhysioNet: A web-based resource for study of physiologic signals

    IEEE Eng. Med. Biol. Mag.

    (2001)
  • G.B. Moody et al.

    ECG data compression for tapeless ambulatory monitors

    Comput. Cardiol.

    (1987)
  • D.A. Lelewer et al.

    Data compression

    ACM Comput. Surv.

    (1987)
  • N.S. Jayant et al.

    Digital Coding of Waveforms

    (1984)
  • K. Sayood

    Introduction to Data Compression

    (2006)
  • A. Gersho

    Principles of quantization

    IEEE Trans. Circuits Syst.

    (1978)
  • A. Buzo et al.

    Speech coding based upon vector quantization

    IEEE Trans. Acoust. Speech Signal Process. ASSP-28

    (1980)
  • R.M. Gray

    Vector quantization

    IEEE Acoust. Speech Signal Process. Mag.

    (1984)
  • M.N. Nasrabadi et al.

    Image coding using vector quantization: a review

    IEEE Trans. Commun.

    (1988)
  • C.M. Kortman

    Redundancy reduction-a practical method of data compression

    Proc. IEEE

    (1967)
  • J.P. Abenstein et al.

    A new data reduction algorithm for real time ECG analysis

    IEEE Trans. Biomed. Eng.

    (1982)
  • S.M.S. Jalaleddine et al.

    ECG data compression techniques-a unified approach

    IEEE Trans. Biomed. Eng.

    (1990)
  • C.A. Andrews et al.

    Adaptive data compression

    Proc. IEEE

    (1967)
  • P. Elias

    Predictive coding

    IEEE Trans. Inform. Theory

    (1955)
  • J.R. Cox et al.

    Compact digital coding of electrocardiographic data

  • U.E. Ruttimann et al.

    Compression of the ECG by prediction or interpolation and entropy encoding

    IEEE Trans. Biomed. Eng.

    (1979)
  • L.W. Gardenhire

    Redundancy reduction, the key to adaptive telemetry

  • L.W. Gardenhire

    Data redundancy reduction for biomedical telemetry

  • Cited by (114)

    • Lightweight electrocardiogram signal compression

      2023, Biomedical Signal Processing and Control
    View all citing articles on Scopus
    View full text