UUnniivveerrssiittyy ooff AArrkkaannssaass,, FFaayyeetttteevviillllee SScchhoollaarrWWoorrkkss@@UUAARRKK Graduate Theses and Dissertations 8-2013 SSyysstteemm DDeessiiggnn AAnndd MMoottiioonn AArrttiiffaacctt RReemmoovvaall AAllggoorriitthhmm IImmpplleemmeennttaattiioonn FFoorr AAmmbbuullaattoorryy WWoommeenn EEccgg MMeeaassuurreemmeenntt SSyysstteemm::ee--BBrraa SSyysstteemm Hyeokjun Kwon University of Arkansas, Fayetteville Follow this and additional works at: https://scholarworks.uark.edu/etd Part of the Cardiovascular Diseases Commons, and the Electrical and Electronics Commons CCiittaattiioonn Kwon, H. (2013). System Design And Motion Artifact Removal Algorithm Implementation For Ambulatory Women Ecg Measurement System:e-Bra System. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/873 This Dissertation is brought to you for free and open access by ScholarWorks@UARK. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of ScholarWorks@UARK. For more information, please contact [email protected]. SYSTEM DESIGN AND MOTION ARTIFACT REMOVAL ALGORITHM IMPLEMENTATION FOR AMBULATORY WOMEN ECG MEASUREMENT SYSTEM : E-BRA SYSTEM SYSTEM DESIGN AND MOTION ARTIFACT REMOVAL ALGORITHM IMPLEMENTATION FOR AMBULATORY WOMEN ECG MEASUREMENT SYSTEM : E-BRA SYSTEM A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Electrical Engineering By Hyeokjun Kwon Hallym University Bachelor of Engineering in Electronics Engineering, 1997 Hallym University Master of Engineering in Electronics Engineering, 1999 August 2013 University of Arkansas This dissertation is approved for recommendation to the Graduate Council. Dr. Vijay K. Varadan Dissertation Director Dr. Randy Brown Dr. Linfen Chen Committee Member Committee Member Dr. Thomas Costello Committee Member ABSTRACT Cardio Vascular Disease (CVD) leads to sudden cardiac death due to irregular phenomenon of the cardiac signal by the abnormal case of blood vessel and cardiac structure. For last three decades, there is an enhanced interest in research for cardiac diseases.. As a result, the death rate by cardiac disease in men has been falling gradually compared with relatively increasing the death rate for women due to CVD. The main reason for this phenomenon is due to the lack of seriousness to female CVD and different symptoms of female CVD compared with the symptoms of male CVD. Usually, because the CVDs for women accompany with ordinary symptoms not attributable to the heart abnormality signal such as unusual fatigue, sleep disturbances, shortness of breath, anxiety, chest discomfort, and indigestion dyspepsia, most women CVD patients do not realize that these symptoms are actually related to the CVD symptoms. Therefore, periodic ECG signal observation is required not only for women who have been diagnosed with heart disease but also for persons who want to examine their heart activity. Electrocardiogram (ECG) is used to diagnose abnormality of heart. Among the medical checkup methods for CVDs, it is very an effective method for the diagnosis of cardiac disease and the early detection of heart abnormality to monitor ECG periodically. This dissertation proposes an effective ECG monitoring system for woman by attaching the system on woman’s brassiere by using augmented chest lead attachment method. The suggested system called “E-Bra system” in this dissertation consists of an ECG transmission system and a computer installed program called “E-Bra pro” in order to display and analyze the ECG transmitted from the transmission module. The ECG transmission module consists of three parts such as ECG physical signal detection part with 3 stage amplifier and two electrodes, data acquisition with AD converter, and data transmission part with GPRS (General Packet Radio Service) communication, and it has very compact size that is attachable at the bottom layer of a brassiere for women. However, the ECG signal measured from the transmission module includes not only pure ECG components information; P waves QRS complex, and T wave, but also a motion artifact component (MA) due to subject movements. The MA component is one of the reasons for misdiagnosis. Therefore, the main purpose of the E-Bra system is to provide a reliable ECG data set identical to the quality of an ECG data set collected in hospital. Unfortunately, removing MA is a big challenge because the frequency range of the MA is duplicated on the frequency range of the pure ECG components, P-QRS-T. In this dissertation, two motion artifact removal algorithms (MARAs) with adaptive filter structure and independent component analysis concept are suggested, and the performance of the two MARAs will be evaluated by correlation values and signal noise ratio (SNR) values. ACKNOWLEDGEMENTS I would like to take this opportunity to express my deepest gratitude to Dr. Vijay K. Varadan, my major advisor and the graduate committee chair, for giving me an opportunity to work under him. He was a fabulous advisor: sharp, cheery, perceptive, and mindful of the things that truly matter. Dr. Randy L. Brown, for his patience, support and guidance in my research and coursework. Dr. Thomas Costello, for his critical reviews, suggestions, constant encouragement and his time. Dr. Linfen Chen for his enthusiasm and making me ponder over engineering applications with civic sense and societal cause. I would like to extend my thanks to my research group members – Dr. Gyanesh Mathur, Prashanth Shyamkumar, Pratyush Rai, Mouli Ramasamy, and Sechang Oh. I was fortunate to have Sunggil Hwang, Sameer Thalappil, and Younglim Choi. My sincerest thanks represents to them for their friendship and support. I offer my sincerest thanks to my parents, brother, and parents-in-law for their support. I am also grateful to sister-in-law family parents for their encouragement. Last but not the least, I would like extend my sincerest thanks to my family, my lovable wife, my princess Soonbi Kwon, and my Jr Taewook Kwon. It would have been very difficult for me to concentrate on my research without their love, affection, and support. TABLE OF CONTENTS Chapter 1: Introduction 1 1.1 Development motivation of electrocardiogram(ECG) measurement system for women 1 1.2 Portable and ambulatory concept for ECG measurement system 6 1.3 Motion artifact removal algorithm requirement 7 1.4 Other studies for removing motion artifact components 10 1.5 Objectives and organization 13 Chapter 2: Fundamental background for ambulatory women ECG measurement system 15 2.1 Fundamental background of electrocardiogram(ECG) 15 2.1.1 Electrocardiogram(ECG) 15 2.1.2 Physiology of the heart 15 2.1.3 Conduction system 16 2.1.4 12 Leads placement for ECG measurement scenarios 17 2.1.4.1 Introduction 17 2.1.4.2 The limb or extremity leads 18 2.1.4.3 The chest or precordial leads 18 2.1.5 ECG waveform features 18 2.1.5.1 ECG rhythm 18 2.1.5.2 P wave 19 2.1.5.3 QRS complex 20 2.1.5.4 T wave 20 2.1.5.5 P-R interval 20 2.1.5.6 Q-T interval 20 2.1.5.7 P-R segment 21 2.1.5.8 S-T segment 21 2.2 Adaptive filter theory 22 2.2.1 Least mean square(LMS) adaptation algorithm 23 2.2.2 Noise cancellation concept by adaptive filter 25 2.3 Independent component analysis(ICA) theory 27 2.3.1 Independent component analysis 27 2.3.2 Entropy 29 2.3.3 FastICA algorithm 31 2.3.3.1 Approximations of negentropy 31 2.3.3.2 Newton method 32 2.3.3.3 FastICA algorithm 33 Chapter 3: E-bra system structure: A women ECG measurement system 35 3.1 Introduction 35 3.2 E-Bra system architecture 37 3.2.1 Hardware structure of the transmission module 40 3.2.1.1 Sensor 41 3.2.1.2 Amplifier 42 3.2.1.3 Microprocessor and power management 43 3.2.1.4 GPRS module 43 3.2.2 Software structure 44 3.2.2.1 Software structure for the transmission module of the E-bra system 44 3.2.2.2 Reliable data transmission guarantee 46 3.2.2.3 Software structure for a remote computer of the E-bra system 48 Chapter 4: Motion artifact removal algorithm architecture 51 4.1 Motion artifact 51 4.2 Noise component extraction algorithm(NCEA) 57 4.2.1 Introduction 57 4.2.2 R-R peak detection 59 4.2.3 R-R Interval synchronization 63 4.3 Simulation of motion artifact removal algorithm 66 4.3.1 SNR measurement 66 4.3.2 Motion artifact removal algorithm with adaptive filter structure 68 4.3.3 Motion artifact removal algorithm with an augmented ICA structure 76 4.3.4 Performance comparison of motion artifact removal algorithm between adaptive filter and ICA 82 Chapter 5: Experimental results of the E-Bra system 89 5.1 Test procedure 89 5.2 Simulation results in the stationary test case with deep breathing 92 5.3 Simulation results in the walking test case 96 5.4 Simulation results in the running test case 100 5.5 Performance evaluation of the MARA with FA Structure and the MARA with ICA sructure for the E-Bra system 103 Chapter 6: Conclusion 106 6.1 Conclusion 106 6.2 Future work 109 REFERENCE 111 APPENDIX 116