Please use this identifier to cite or link to this item:
Title: ECG Signal Compression for Transmission in Mobile Health Monitoring System
Authors: Kaur, Arshdeep
Supervisor: Joshi, Hem Dutt
Keywords: ECG;Compression;Wavlet;Electronics and Communication;ECED
Issue Date: 30-Oct-2015
Abstract: The Electrocardiogram signal (ECG) is the most used biomedical signal giving most accurate analysis for heart/cardiovascular disease (CVD). CVD is the number one cause of deaths worldwide causing 17.5 million deaths in 2012 which amounts to 31% of total number of deaths. Heart failure is a huge leverage on society because of its high costs of treatment, lower quality of life and untimely death. Statistics show that more than half of health agencies in US use telemonitoring services, where ECG signals are transmitted over mobile networks for continuous monitoring of patients from a distance. This is called mHealth i.e. use of mobile communication in medical industry. It is not only financially advantageous but also gives patient freedom to live a normal life. However this data is enormously huge and it is not economically feasible to transmit the data in raw form. ECG records time varying electrical pulses generated in the heart and is primarily tool for evaluating and identifying cardiac disorders. It is sampled between 100 - 1000 Hz at 8–16 bit resolution. The data rate is 11–22 Mbits/h/lead approximately which is huge. So ECG data cannot be sent in raw format over wireless channels. Data needs to be compressed for transmission. This report discusses ECG compression using wavelet transforms out of all available techniques as they have shown to be promising contender for the proposed objective. As a part of this thesis report, an existing compression algorithm proposed by Bashar A. Rajoub has been implemented and its performance is compared with the new algorithm proposed in this report. In this report a novel approach of empirical wavelet based ECG compression using set partitioning in hierarchical trees algorithm (SPIHT) for mobile device based application has been presented. The proposed algorithm implements empirical wavelet transform with SPIHT encoding. SPIHT is an embedded computationally simple coding algorithm suitable for wireless transmission. This algorithm was applied on different datasets of MIT-BIH database and the results show high accuracy and compression ratio when compared with existing ECG compression techniques. From the simulation it has been found that best performance of 35.9:1 CR with 1.15% PRD at 160bps for record 117 is achieved.
Description: M.E. (Wireless Communications)
Appears in Collections:Masters Theses@ECED

Files in This Item:
File Description SizeFormat 
3838.pdf1.51 MBAdobe PDFThumbnail

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.