Lossless electrocardiogram signal compression: A review of existing methods. (May 2019)

• Record Type: Journal Article
• Authors: Tiwari, Abhishek; Falk, Tiago H.
• Is Part Of: Biomedical signal processing and control. Volume 51(2019)
• Page Start: 338
• Page End: 346
• Publication Date: 2019-05
• Subjects: Electrocardiogram -- ECG -- Lossless Compression -- Review; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2019.03.004
• Languages: English
• ISSNs: 1746-8094

Abstract

Abstract: Background: Cardiovascular diseases (CVDs) are among one of the leading causes of death in the world today. Electrocardiography (ECG) is commonly used to monitor and diagnose heart disorders at an early stage. In recent years, with the burgeoning of wearable technologies, portable ECGs have been applied not only for tele-health cardiac monitoring and diagnostic applications, but also for stress monitoring, fitness analysis, and general health assessment, to name a few. Such devices are capable of generating a wealth of data, but due to transmission and storage limitations, the majority of the recorded ECG data is discarded and lower-dimensional features, such as heart rate or heart rate variability, are stored instead. Devices that do allow for ECG recording/streaming, in turn, typically apply lossy compression algorithms, thus are not applicable for clinical use. While lossless compression techniques have been widely used in allied domains, limited application has been seen for ECGs. Objective: This literature review aims at providing the research community a summary of lossless compression methods developed specifically for ECGs and compares existing methods based on the depth and breadth of the databases in which they were tested, the specific compression algorithms used, and how their performances were evaluated. Methods: English peer-reviewed journal articles published between 1990 and 2017 were chosen as the target of this review. A data extraction sheet was then prepared to group and categorize articles to developing a better understanding of the research directions taken by the community and the important components to be taken into account while working with ECG lossless compression algorithms. Results: The different articles were grouped and analyzed on the basis of the databases used, pre-processing, compression methods, type of implementation, performance measures used and comparisons with prior art. Several recommendations and research directions were provided based on the current work. Conclusion: It is hoped that this review will provide technology developers with invaluable insights, thus opening doors for wearables devices with clinically-relevant capabilities.