Undergraduation Thesis Topic Personal Authentication By Single-Channel Ecg Signals.docx

HANOI UNIVERSITY OF SCIENCE AND TECHNOLOGY SCHOOL OF ELECTRONICS AND TELECOMMUNICATIONS UNDERGRADUATION THESIS Topic PERSONAL AUTHENTICATION BY SINGLE CHANNEL ECG SIGNALS Student VU THI MINH Biomedica[.]

HANOI UNIVERSITY OF SCIENCE AND TECHNOLOGY

SCHOOL OF ELECTRONICS AND TELECOMMUNICATIONS

UNDERGRADUATION THESIS

Topic:

PERSONAL AUTHENTICATION BY

SINGLE-CHANNEL ECG SIGNALS

Student: VU THI MINH

Biomedical Engineering ClassAdvanced Program – Course 58Supervisor: Dr NGUYEN VIET DUNGArgue Officer:

Hanoi, 1-2019

Thesis Assessment (For Supervisor)

Supervisor: Nguyen Viet Dung

Subject: Personal authentication by single- channel ECG signals.

Choose the appropriate scores for students following the criteria below: Very poor (1); Poor (2); Average (3); Good (4); Excellent (5)

The combination of theory and practice (20)

1 Clearly state the urgency and importance of the subject, theproblems and assumptions (include purposes and relevance) as well

2 Update the latest research results (national/international) 1 2 3 4 5

3 Clearly state the study/problem solving methodology in detail 1 2 3 4 5

4 Have stimulation/experimental results and describe obtained resultsin detail. 1 2 3 4 5

Ability to analyze and evaluate the results (15)

5 Clear working plan, including objectives and methodology based onsystemically theoretical study results. 1 2 3 4 5

6 Results are presented logically and easy to understand; all resultsare satisfactorily analyzed and assessed. 1 2 3 4 5

7 In conclusion section, author specifies the differences (if any)between the results obtained and the initial objectives while

providing arguments to propose possible solutions in the future 1 2 3 4 5

Writing skill (10)

8

The thesis is presented on a prescribed format with logical and nice

structure of chapters (Tables, F

igures are clear with captions, are numbered and explained or

mentioned in the thesis; has alignments, has spaces after full stops

and commas, etc.), has chapter introductions and conclusions, listed

references and citations following regulations

1 2 3 4 5

9 Excellent writing skill (right syntax, scientific style, logicalreasoning, appropriate vocabularies, etc.). 1 2 3 4 5

Science research achievements (5) (choose 1 in 3 options)

10a Had the published or accepted scientific articles/3rd prize atSchool level at student science research conference or higher/3rd

scientific prize (international/national) or higher/registered patents 5

10b

Reported at School-level board in student science research

conference but not achieved 3rd prize or higher/Achieved

consolation prize at other nationally or internationally specialized

competitions such as TI contest

3 Other comments of instructor (Instructor comments on student’s work

attitude and spirit)

Date: …./… /2019

Supervisor (Signature & full name)

Thesis Assessment (For Argue Officer)

Argue Officer: ………

Subject: Personal authentication by single- channel ECG signals.

Choose the appropriate scores for students following the criteria below: Very poor (1); Poor (2); Average (3); Good (4); Excellent (5)

The combination of theory and practice (20)

1 Clearly state the urgency and importance of the subject, theproblems and assumptions (include purposes and relevance) as well

2 Update the latest research results (national/international) 1 2 3 4 5

3 Clearly state the study/problem solving methodology in detail 1 2 3 4 5

4 Have stimulation/experimental results and describe obtained resultsin detail. 1 2 3 4 5

Ability to analyze and evaluate the results (15)

5 Clear working plan, including objectives and methodology based onsystemically theoretical study results. 1 2 3 4 5

6 Results are presented logically and easy to understand; all resultsare satisfactorily analyzed and assessed. 1 2 3 4 5

7 In conclusion section, author specifies the differences (if any)between the results obtained and the initial objectives while

providing arguments to propose possible solutions in the future 1 2 3 4 5

Writing skill (10)

8

The thesis is presented on a prescribed format with logical and nice

structure of chapters (Tables, Figures are clear with captions, are

numbered and explained or mentioned in the thesis; has alignments,

has spaces after full stops and commas, etc.), has chapter

introductions and conclusions, listed references and citations

following regulations

1 2 3 4 5

9 Excellent writing skill (right syntax, scientific style, logicalreasoning, appropriate vocabularies, etc.). 1 2 3 4 5

Science research achievements (5) (choose 1 in 3 options)

10a Had the published or accepted scientific articles/3rd prize atSchool level at student science research conference or higher/3rd

scientific prize (international/national) or higher/registered patents 5

10b

Reported at School-level board in student science research

conference but not achieved 3rd prize or higher/Achieved

consolation prize at other nationally or internationally specialized

competitions such as TI contest

2

Total score on scale of 10

3 Other comments of argue officer (Argue officer comments on

student’s work attitude and spirit)

Date: …./… /2019 Argue officer (Signature & full name)

Biometrics refers to the recognition of individuals based on physiological orbehavioral characteristics Biometric traits include the retina, face, iris, fingerprints,and voice Through these information sources, various methods can recognizeindividuals However, some limitations of these technologies lead to require highersecurity The electrocardiogram (ECG) is one of the most commonly knownbiological signals ECG involves information about the structural and functionalcardiac muscle activities, and it is a simple and effective representative of anoninvasive diagnostic method Every individual has characteristic ECG features.Therefore, such signals provide strong protection against forgery Recently, moreresearch has focused on extracting ECG features with different method in availabledatabase have high accuracies In the thesis, a goal is to extract more suitablefeatures with signals, which are acquired from Kardia Mobile device, thus,identifying individuals

I wish to express my sincere thanks to my instructor Dr Nguyen Viet Dungfor his devoted guidance and supervision This practice would not have beencompleted without his care and dedication in constructively criticizing my work Ialso wish to express my sincere thanks to the lecturers in School of Electronics andTelecommunications as well as in Hanoi University of Science and Technologywho have taught me countless useful knowledge Finally, I special thanks goes to

my parents and other family members, who have unconditionally given me all oftheir support on all fronts They inspired and gave me strengths to finish this work

Hanoi, January 2th 2019

Student

Vu Thi Minh

Biosignals contain useful information that can be used to identify individualsbeside applications in medical Biological signals can be classified according tovarious characteristics of the signal, including the waveform shape, statisticalstructure, and temporal properties Biosignals can prevent to falsify from physicalfeatures in biometrics such as face, fingerprint, iris, etc An Electrocardiogram(ECG) measures and records the electrical activity that passes through the heart Inthis study, I researched single- channel electrocardiogram (ECG) signal which is gotfrom a device named Kardia mobile designed by AliverCor company and hasmedical standards certification from FDA A feature set extracted based on theassociation between Discrete Wavelet Transform (DWT) and Statistic data waspropounded and Support Vector Machine (SVM) was exerted for ECGclassification Results show that my method achieves approximately 87.5% for datathat I collected However, the amount of data used for training is limited

Tín hiệu sinh học chứa những thông tin hữu ích có thể được sử dụng để xácđịnh các cá nhân bên cạnh các ứng dụng trong y tế Tín hiệu sinh học có thể đượcphân loại theo các đặc tính khác nhau của tín hiệu, bao gồm dạng sóng, cấu trúcthống kê và tính chất thời gian Tín hiệu sinh học có thể ngăn ngừa việc làm giả cácđặc điểm vật lý trong sinh trắc học như khuôn mặt, dấu vân tay, mống mắt, v.v Điện tâm đồ (ECG) đo và ghi lại hoạt động điện đi qua tim Trong nghiên cứu này,tôi đã nghiên cứu tín hiệu điện tim đơn kênh (ECG) được lấy từ một thiết bị có tênKardia mobile do công ty AliverCor thiết kế và có chứng nhận tiêu chuẩn y tế từFDA Một bộ tính năng được trích xuất dựa trên sự liên kết giữa biến đổi Waveletrời rạc (DWT) và dữ liệu thống kê đã được đưa ra và máy vector hỗ trợ (SVM) đãđược sử dụng để phân loại ECG Kết quả cho thấy phương pháp của tôi đạt đượcxấp xỉ 87.5% đối với dữ liệu mà tôi đã thu thập Tuy nhiên lượng dữ liệu được sửdụng bị hạn chế

ACKNOWLEDGMENTS 6

ABSTRACT 7

LIST OF FIGURES 11

LIST OF ACRONYMS 14

CHAPTER 1: FUNDAMENTAL BACKGROUND INFORMATION 16

1.1 Biometrics authentication 16

1.1.1 Types of biometrics 16

1.1.2 Basic components of biometric system 25

1.1.3 Some criteria of biometrics 25

1.2 Electrocardiography (ECG/ EKG) 26

1.2.1 Definition 26

1.2.2 ECG waveform 27

1.2.3 Different noise in ECG signals 31

1.3 Wavelet Transform (WT) and Discrete Wavelet Transform (DWT) 33

1.3.1 Fundamental Concepts and Overview of Wavelet Transform 33

1.3.2 Multi-resolution Analysis and Continuous Wavelet Transform 38

1.3.3 Multi-resolution Analysis: Discrete Wavelet Transform 43

1.3.4 Wavelet families 50

1.4 Statistics data 58

1.5 Machine learning (ML) 59

1.5.1 Types of machine learning 59

1.5.2 Support Vector Machine (SVM) 61

1.5.3 K- nearest neighbor (KNN) 69

CHAPTER 2 SIGNAL PREPARATION 74

2.1 ECG Acquisition 74

2.1.1 ECG Recording equipment: Kardia mobile 74

2.1.2 Web plot digitizer 76

2.2 Experiment set up 78

CHAPTER 3: DESIGNING ECG BASED PERSONAL AUTHENTICATION SYSTEM 81

3.1 Block diagram 81

3.1.1 Pre-processing 81

3.1.2 Feature extraction Algorithm from DWT using Daubechies wavelet 83

3.1.3 Classification 86

3.2 Results and Discussion 87

CONCLUSION 91

LIST OF FIGURES

Figure 1.1 Types of biometrics 17

Figure 1.2 Retina scanning 17

Figure 1.3 Iris sample 19

Figure 1.4 Fingertip 20

Figure 1.5.Fingerprint matching mechanism 20

Figure 1.6 Automatic face recognition system 21

Figure 1.7 Sample voice clip as shown in sound editor 22

Figure 1.8 Hand/Palm-print and hand/palm-print features 24

Figure 1.9 Flow chat of a biometrics 25

Figure 1.10 Schematic anatomy of the human heart 27

Figure 1.11 Typical cardiac waveform 27

Figure 1.12 Normal sinus rhythm [21] 29

Figure 1 13 Sinus Bradycardia [21] 29

Figure 1 14 Sinus Tachycardia [21] 30

Figure 1.15 Schematic representation of various QRS complex configurations and normal ranges of wave amplitudes and durations [23] 30

Figure 1.16 The classical ECG curve with its most common waveform 31

Figure 1.17 60 Hz Power Line Interference[25] 32

Figure 1.18 ECG signal which contains EMG noise [26] 32

Figure 1.19 Sine wave with frequencies at 3Hz, 10 Hz and 50 Hz [29] 34

Figure 1.20 The FT of 50Hz signal [29] 35

Figure 1.21 Example of a signal has high frequency components for short 38

Figure 1.22 Cosine signals corresponding to various scale [29] 41

Figure 1.23 Illustration of the entire process CWT [29] 43

Figure 1.24 The subband coding Algorithm 43

Figure 1.25 The scale function and mother wavelet function of Haar wavelet 51

Figure 1.26 Meyer Scaling function 52

Figure 1.27 Fourier transform of the scaling function of the Meyer wavelet 52

Figure 1.28 Fourier transform of Meyer wavelet 53

Figure 1.29 Meyer wavelet 54

Figure 1.30 Real part of the Morlet wavelet for ωo = 5 55

Figure 1.31 Imaginary part of the Morlet wavelet for ωo = 5 58

Figure 1.32 The Daubechies scaling function for N = 4 57

Figure 1.33 The Daubechies wavelet for N = 4 58

Figure 1.34 ML techniques 60

Figure 1.35 Types of algorithm 61

Figure 1.36 Example of the training data 62

Figure 1.37 Example of the separating hyperplane 63

Figure 1.38 Several separating hyperplanes 63

Figure 1.39 Margin 64

Figure 1.39 Analysis of the SVM problem .65

Figure 1.40 The point’s closest to separable plane of the two classes is circled 67

Figure 1.41 Soft SVM margin a) Noise data, b) Near linearly separable data 68

Figure 1.42 Introduce slack variables ξn 68

Figure 1.43 Map of 1NN 70

Figure 2.1 The Kardia mobile 75

Figure 2.2 Kardia software 76

Figure 2.3 Obtained ECG from Kadia 77

Figure 2.4 WebPlot digitizer 78

Figure 2.5 Putting fingers on device 80

Figure 3.1 Raw signal after digitizer 81

Figure 3.2 Histogram of origin signal (above) and synthesis signal (below) 82

Figure 3.3 Daubechies 4 wavelet 84

Figure 3.4 Daubechies 6 wavelet 84

Figure3.5.Frequency components of each decomposition level and the corrresponding frequency bands .85

Figure 3.6 Signal and power spectrum before and after filtering 87

Figure 3.7 Reconstructed signal level 1, 2, 3, 4 88

LIST OF TABLES

Table 3.1 Investigation the accuracy with rations test / train 89Table 3.2 Accuracy of training and testing 89Table 3.3 The actual accuracy, FAR, FRR in survey 90

LIST OF ACRONYMS

CO Correlation distance metric

CHAPTER 1: FUNDAMENTAL BACKGROUND

INFORMATION1.1 Biometrics authentication

Biometrics is a technology used to identify, analyze, and measure an individual’s

physical and behavioral characteristics A biometric system is a technology, whichtakes an individual physiological, behavioral, analyzes it, and identifies theindividual as a genuine or malicious user Biometric identification consists ofdetermining the identity of a person The aim is to capture an item of biometric datafrom this person It can be a photo of their face, a record of their voice, or an image

of their fingerprint This data is then compared to the biometric data of several otherpersons kept in a database In this mode, the question is a simple one: "Who areyou?" Biometric authentication is the process of comparing data for the person'scharacteristics to that person's biometric "template" in order to determineresemblance The reference model is first store in a database or a secure portableelement like a smart card The data stored is then compared to the person'sbiometric data to be authenticated Here it is the person's identity which is beingverified In this mode, the question being asked is: "Are you indeed Mr or Mrs X?"[1]

Personal biometric authentication: In this problem, I just let one person enter the

system, so I just need to identify the person who are you or not

Overall biometric authentication: The problem will allow many people who had

database in the system to enter it

In my project, I solve the Personal biometric authentication problem I indicate thesignal that is of A, or not A person

1.1.1 Types of biometrics

Biometric devices are many types, but majorly there are five types of biometricssecurity which are commonly used Biometrics is basically the recognition ofhuman being personality that are unique to each human, which includes facialrecognition, fingerprints, voice recognition, retina scans, palm prints, and more has

shown in Figure 1.1 Biometric technology are used to keep the devices safe in thebest way to ensure that people stay out of their valuable assets and information, andwill find that using any one of these five biometrics security, devices is a great way

to keep things safe.[2]

Figure 1.1 Types of biometrics.

o Retina scanner

Retina scanning is a biometric verification technology that uses an image of anindividual’s retinal blood vessel pattern as a unique identifying trait for access tosecure installations The human retina is a thin tissue composed of neural cells that

is located within the posterior part of the eye as shown in Figure 1.2

Figure 1.2 Retina scanning

Due to the complex shape of the capillaries that deliver the retina with blood, alland sundry's retina is unique The network of blood vessels within the retina is socomplicated that even identical twins do not proportion a comparable sample Eventhough retinal styles can be altered in instance of diabetes, Glaucoma or retinaldegenerative disorders, the retina typically remains unaffected from birth till dying.Due to its unique and unchanging nature, the retina seems to be the maximumprecise and dependable biometric [3].

o Iris Scanning

Iris recognition uses digital camera technology, with slight infrared illuminationlowering specular reflection from the convex cornea, to create photographs of thedetail-wealthy, elaborate systems of the iris as shown in Figure 1.3 Converted intodigital templates, those snap shots offer mathematical representations of the iris thatyield unambiguous wonderful identity of an individual Iris reputation efficiency isnot often impeded by using glasses or contact lenses Iris technology has thesmallest outlier (folks that cannot use/enroll) group of all biometric technologies[4] Moreover, it has a small template size that allows speedy comparisons makingiris recognition technology particularly well suited for one-to-many identifications.Even genetically identical individuals have distinct iris textures which furtherconfirm that it is a highly accurate and reliable technique Because of its pace ofcontrast, iris reputation is the handiest biometric technology nicely perfect for one-to-many identity Advantage of iris reputation is its balance, or template sturdiness,

a single enrollment can closing an entire life There are few benefits of the use ofiris as biometric identification: it's far an inner organ this is properly includedagainst damage and wear by a rather obvious and touchy membrane (the cornea)[5]

A new technology requires substantial investment and hence may not be suitablefor small organizations It is quite difficult to perform iris recognition from adistance larger than a few meters and the subject to be identified needs to be co-operative The subject should hold his or her head still and look into the camera Irisrecognition is also susceptible to poor quality of images as well as associated failure

When it comes to biometrics, the iris has several major advantages compared to afingerprint:

- You do not spread the information around every time you touch something

- The iris stays virtually unchanged throughout a person’s life A fingerprint,

on the other hand, can be dirtied, scarred or eroded

- You cannot use a fingerprint with dirty or sweaty hands Irises, however,have no such problem.[6]

However, both retina scanners and iris scanners have proven to be easy to tricksimply by using a high-quality photograph of the subject’s eyes or face

Figure 1.3 Iris sample.

o Finger print scanner

Fingerprints are the graphical glide-like ridges gift on human palms Finger ridgeconfigurations do no longer exchange for the duration of the life of a person besidesdue to accidents including bruises and cuts on the fingertips This belongings makesfingerprints a totally attractive biometric identifier Fingerprint-based (Figure 1.4)totally private identification has been used for a very long time As a long way asfee is going, the fingerprint scanning is on the lower stop of the dimensions Themost inexpensive fingerprint scanners are those that best scan the actual print,though the dearer ones really experiment the presence of blood in the fingerprint,the scale and shape of the thumb, and plenty of different features as shown in Figure

1.5 Those costlier structures in reality capture a 3D photo of the fingerprint,thereby making it a great deal more difficult for the fingerprint to be counterfeited[7]

Figure 1.4 Fingertip

Figure 1.5.Fingerprint matching mechanism

Tests conducted by the International Biometric Group on fingerprint systems ofparticipating vendors found that the false acceptance rate of these systems rangedfrom 0% to 5% On the same day of enrolment, tests were conducted for falserejection rate and it ranged from 0% to 35% However, when the tests wereconducted six weeks later the FRR’s ranged from 0% to 66% Systems from somevendors worked very well while others had some accuracy problems When vendor-independent tests were conducted by the FVG2006, it found that FAR was held

constant at a rate of 0.01% This rate is considered sufficient for most authenticationscenarios.

o Facial biometrics

Each individual has a distinctly unique face, even two twins that cannotdifferentiate by human’s eye There are sure markers that enable these biometricacknowledgment scanners to in a split second recognize the uniqueness of everyindividual examining their facial elements Face Recognition System (Figure 1.6)measures and matches the unique characteristics for the purposes of identification orauthentication Often leveraging a digital or connected camera, facial recognitionsoftware can detect faces in images, quantify their features, and then match themagainst stored templates in a database

Figure 1.6 Automatic face recognition system.

Several factors can affect the accuracy of facial biometrics It might not workwell under poor lighting conditions, the presence of sunglasses or other objects thatpartially cover the subject’s face and low resolution images Also the face of aperson changes over time The accuracy of some facial recognition systems is alsoaffected by the variation in facial expressions and for this reason most countriesallow only neutral facial expressions in passport photos

A real-world facial biometrics accuracy test verified passport photos ofpassengers against a lesser quality live scan photo taken within the border controlgates itself The top performing vendor in the National Institute of StandardsTechnology (NIST) test achieved a FRR of 1.1% Thus, face recognition systemsdefinitely provide better accuracy when compared to live guards performing amanual comparison of passport photos with the passport holders.

o Voice recognition

Voice or speaker recognition is the ability of a machine or program to receiveand interpret dictation or to understand and carry out spoken commands Voicerecognition has gained prominence and use with the rise of AI and intelligentassistants, such as Amazon's Alexa, Apple's Siri and Microsoft's Cortana Eachperson in the world has a unique voice pattern as shown in Figure 1.7, even thoughthe changes are slight and hardly noticeable to the human ear On the other handwith uncommon voice recognition programming, those moment contrasts in everyindividual's voice can be noted, experienced and validated to enable the access tothe individual that has quality pitch which is a correct one, and at the same timevoice level also

Figure 1.7 Sample voice clip as shown in sound editor.

Voice recognition systems enable consumers to interact with technology simply

by speaking to it, enabling hands-free requests, reminders and other simple tasks.Surprisingly it can be effective at differentiating two people who have almost

identical voice patterns Voice recognition software on computers requires thatanalog audio be converted into digital signals, known as analog-to-digitalconversion For a computer to decipher a signal, it must have a digital database, orvocabulary, of words or syllables, as well as a speedy means for comparing this data

to signals The speech patterns are stored on the hard drive and loaded into memorywhen the program is run A comparator checks these stored patterns against theoutput of the A/D converter an action called pattern recognition

Background noise can produce false input, which can be avoided by using thesystem in a quiet room There is also a problem with words that sound alike, butthat are spelled differently and have different meanings for example, hear andhere Feeding the wrong voice cannot always be avoided in voice recognition aswell as the voice capturing machine should be near to the user

o Palm vein pattern recognition

Palm vein is one of the most secure biometrics and is the world’s firstcontactless personal identification system It works by capturing the vein patternimage of an individual while radiating it with near-infrared rays Its speciality is that

it can detect the vein pattern on the human palm with utmost precision When thesensor emits a near-infrared ray towards the palm of the hand, the blood flowingthrough these back to the heart with reduced oxygen absorbs the radiation andcauses the veins to appear as a black pattern This pattern is then recorded andstored in an encrypted format in a database, token or smart card as a reference forfuture comparison

By placing your hand on a scanner, you not only have a unique fingerprintpattern, but the size and shape of your entire hand is also very unique as shown inFigure 1.8 It differs to a unique finger impression in that it likewise contains otherdata, for example, touch, indents and symbols which can be utilized whencontrasting one palm with another Hand prints can be used for criminal, forensic orcommercial applications [8] The main difficulty of hand print is that the printchanges with time depending on the type of work the person is doing for anextended duration of time

Figure 1.8 Hand/Palm-print and hand/palm-print features.

It uses the information that is contained within a person’s body to confirm his orher identity Therefore it is highly accurate because the vein pattern of the humanpalm is not only complex but it is also unique to each individual It uses acombination of image recognition and optical technology to scan the normallyinvisible vein pattern of the human palm This makes it highly resistant to anyspoofing methods such as impersonation, counterfeiting or any other dishonestactions Moreover, it is designed in such a way that it can detect only the veinpattern of living persons It also has a fast scanning process and does not need anycontact which means that this technology meets the strict hygiene requirements thatare usually required for usage in public environments [9]

o Live-biosignal authentication

Besides the popular authentication methods listed above, bio-signal alsoresearched to recognize individuals because of their features such asElectromyogram (EMG) and Electroencephalogram (EEG)

Electromyography (EMG) is known as process of recording the electricactivities of muscles This signal is Electromyogram Electromyogram is an activitypotential signal generated when muscle contracts, which is controlled by thenervous system and produced during muscle contraction The signal represents theanatomical and physiological properties of muscles [10]

The electroencephalogram (EEG) signal represents a subject’s specific brainactivity patterns and is considered as an ideal biometric given its superior forgeryprevention However, the accuracy and stability of the current EEG-based personauthentication systems are still unsatisfactory in practical application However, theaccuracy and stability of the current EEG-based person authentication systems arestill unsatisfactory in practical application [11].

Based on different individual features, they can distinguish with otherindividuals However, both EMG and ECG signals are just only used in medical,not widely popular

1.1.2 Basic components of biometric system

There are four general steps in a biometric system taking to performidentification and verification as shown in Figure 1.9

 Acquire live sample from candidate (Use sensors)

 Extract prominent features from sample (Use processing unit)

 Compare live sample with samples stored in database (Use algorithms)

 Present the decision (Accept or reject the candidate)

Figure 1.9 Flow chat of a biometrics 1.1.3 Some criteria of biometrics

False Acceptance Rate (FAR) – It is the measure of possibility that a biometric

system will incorrectly identify an unauthorized user as a valid user [12]

FAR = Number of Identification Attemps Number of False Acceptances

A biometric system providing low FAR ensures high security

(1.1)

False Reject Rate (FRR) – It is the measure of possibility that the biometric

system will incorrectly reject an authorized user as an invalid user [12]

FRR = Number of Identification Attemps Number of False Rejections

If FAR is equals to FRR, the obtained results are the best

1.2 Electrocardiography (ECG/ EKG)

The electrocardiogram (ECG), resulting from the electrical conduction throughthe heart needed for its contraction, is one of the most recent traits to be exploredfor biometric purposes [13, 14] Despite being far from as developed or widespread

as face or fingerprint biometrics, the ECG offers unique advantages in terms ofuniversality, uniqueness, permanence, and liveness assurance, that attest itspotential for the recognition of individuals [15,16]

1.2.1 Definition

An EKG, also called an ECG or electrocardiogram, is a recording of the heart'selectrical activity It is a quick and painless procedure EKGs captures a tracing ofcardiac electrical impulse as it moves from the atrium to the ventricles Theseelectrical impulses cause the heart to contract and pump blood [17]

The leads are placed on specific locations of the body of the person to recordECG either on graph paper or on monitors The human heart contains four chambersi.e., Right Atrium, Left Atrium, Right Ventricle and Left Ventricle The upperchambers are the two Atria’s and the lower chambers are the two Ventricles Underhealthy condition, the heartbeat begins at the Right Atrium called Sino Atria (SA)node and a special group of cells send these electrical signals across the heart Thissignal travels from the Atria to the Atrio Ventricular (AV) node The AV nodeconnects to a group of fibers in Ventricles that conducts the electrical signaland transmits the impulse to all parts of the lower chamber, the Ventricles Toensure that the heart is functioning properly this path of propagation must betraced accurately [18].The basic structure of heart is depicted in Figure 1.10

(1.2)

Figure 1.10 Schematic anatomy of the human heart.

To determine the potential use of ECG as a biometric, it is necessary to evaluatehow ECG satisfies the requirements for biometric characteristics

A "perfect" biometric characteristic should be:

o Universal, i.e., each individual possesses this characteristic

o Easily measured, i.e., it is quite easy technically and convenient for anindividual to obtain the characteristic

o Unique, i.e., there are no two individuals with identical characteristics

o Permanent, i.e., the characteristic does not change over time

"Good" biometric characteristics can to a greater or lesser extent satisfy theserequirements, depending on the purpose and application of biometric system

1.2.2 ECG waveform

Each heart beat displayed is a sequence of electrical waves characterized bypeaks and valleys ECG mainly provides two kinds of information One is theduration of the electrical wave passing through the heart and it will decide whetherthe electrical activity is normal ,or slow, or irregular Second is the amount ofelectrical activity passing through the heart muscle that helps to find whether theparts of the heart are too large or overworked The frequency range of an ECGsignal is 0.05– 100 Hz and its dynamic range is 1–10 mV The ECG signal ischaracterized by five peaks and valleys represented by the letters P, Q, R, S, T.Sometimes U wave is also present The performance of ECG analysis is based onthe accurate and reliable detection of the QRS complex as well as T- and P waves[19] [20] An ideal ECG wave is as shown in Figure 1.11:

Figure 1.11 Typical cardiac waveform

o P waves are the wave of atrial depolarization ( amplitude A< 0.25 mV, time

<0.1s)

o T waves are atrial repolarization that is not seen on the ECG because

ventricular depolarization is stronger

o QRS waves are ventricular depolarization state (total time: 0.08s)

- Q wave A 0.3 mV, time 0.03s (begin ventricular depolarization

process)

- R-wave A <= 2mV (depolarization simultaneously both ventricles)

- S wave: bottom polar ventricular depolarization (as Q wave)

o ST, T and U : repolarization of the ventricles ( amplitude < 0.5mV,

time =0.2s)

The P-wave represent the activation of the two atria, the upper chambers of theheart, while the QRS complex and T-wave represent the excitation of the lowerchamber of the heart, the ventricles QRS detection is one of the fundamental issues

in automatic ECG signal analysis After QRS complex has been detected a thoroughexamination of ECG signal is done The P, QRS and T-waves reflect therhythmic electrical depolarization and repolarization of the myocardium linkedwith the contractions of the atria and ventricles [21] The horizontal section

of this waveform prior to the P-wave is termed as the baseline or theisopotential line The P-wave corresponds to the depolarization of the atrialmusculature The QRS complex gives the combined result of the repolarization ofthe atria and depolarization of the ventricles, which occurs almost at same time TheT-wave is the wave of ventricular repolarization, whereas the U-wave, if present isnormally believed to be the result of after potentials in the ventricular muscle

So the duration amplitude and morphology of the QRS complex is helpful in

diagnosing cardiac arrhythmias, conduction abnormalities, ventricularhypertrophy, myocardial infection and other disease states The usual rate ofheart is 60 to 100 beats per minute A slower rate than the normal range is calledbradycardia (slow heart) and a higher rate is called tachycardia (fast heart) If theECG signal is not normal then an Arrhythmia is indicated [21, 22] The waveform

of normal ECG waveform and that of the abnormalities is shown in Figures1.12, 1.13 and 1.14:

Figure 1.12 Normal sinus rhythm [21].

Figure 1.13 Sinus Bradycardia [21].

Figure 1.14 Sinus Tachycardia [21]

It is plausible to assume that an ECG is an almost unique human characteristicbecause morphology and amplitudes of recorded cardiac complexes are governed

by multiple individual factors, in particular by the shape and position of the heart,and the presence and nature of pathologies, among other factors As a result, QRScomplexes have a variety of configurations and metrics (Figure 1.15 and Figure1.16)

Figure 1.15 Schematic representation of various QRS complex configurations and normal ranges

of wave amplitudes and durations [23].

Figure 1.16 The classical ECG curve with its most common waveform.

1.2.3 Different noise in ECG signals

Different noises are affected by the ECG signal during its acquisition andtransmission Mainly two types of noises are present in the ECG signal Noises withhigh frequency include Electromyogram noise, Additive white Gaussian noise, andpower line interference Noises with low frequency include baseline wandering Thenoises contaminated in the ECG signal may lead wrong interpretation There aredifferent denoising techniques are available in the literature [24]

o Power line interference

It consist of 50/60Hz pickup and harmonics The interference is mainly caused

by Electromagnetic interference by power line, Electromagnetic field (EMF) by thenearby machines, Stray effect of the alternating current fields due to loops in thecables, Improper grounding of patient or the ECG machine ,The electricalequipment induce 50/60 Hz signals in the input circuits of the ECG machine.Example, air conditioner, elevators and X-ray draw heavy power, line current [25]

Figure 1.17 60 Hz Power Line Interference[25]

o Electromyogram (EMG) Noise

The Electromyogram (EMG) noise is generated from electrical activity of themuscle EMG consist of maximum frequency of 10 KHz Sections of ECG may beinterfered and corrupted by surface EMG which causes difficulties in dataprocessing and analysis[26]

Figure 1.18 ECG signal which contains EMG noise [26].

o Baseline Wander

Baseline wander is a low-frequency noise component present in the ECG signal.This is mainly due to respiration, and body movement The frequency content ofbaseline wander is usually in the range below 0.5 Hz; however, increasedmovement of the body during the latter stages of a stress test further increases thefrequency content of baseline wander [27] This low frequency noise, Baselinewander causes problem in detection and analysis of peak

o Channel Noise

Channel noise introduces when ECG signal is transmitted through channels This

is due to the Poor channel conditions It is mainly like white Gaussian noise, whichcontains all frequency components [25] E.g AWGN

o Electrode Contact Noise

Electrode contact noise is caused by the loss of contact between the electrode andthe skin, which effectively disconnects the measurement system from the subject.The noise is of duration 1s [28]

o Motion artifacts

Motion artifacts are transient base line changes caused by changes in theelectrode-skin impedance with electrode motion As this impedance changes, theECG amplifier sees a different source impedance which forms a voltage dividerwith the amplifier input impedance therefore the amplifier input voltage dependsupon the source impedance which changes as the electrode position changes [28]

1.3 Wavelet Transform (WT) and Discrete Wavelet Transform (DWT)

1.3.1 Fundamental Concepts and Overview of Wavelet Transform

Mathematical transformations are applied to signals to obtain a furtherinformation from that signal that is not readily available in the raw signal There arenumber of transformations that can be applied, among which the Fourier transformsare probably by far the most popular Most of the signals in practice, are time-domain signals in their raw format That is, whatever that signal is measuring, is afunction of time In other words, when we plot the signal one of the axes is time(independent variable), and the other (dependent variable) is usually the amplitude.When we plot time-domain signals, we obtain a time-amplitude representation ofthe signal This representation is not always the best representation of the signal formost signal processing related applications In many cases, the most distinguishedinformation is hidden in the frequency content of the signal The frequencyspectrum of a signal is basically the frequency components (spectral components) of

that signal The frequency spectrum of a signal shows what frequencies exist in thesignal [29].

The frequency is measured in cycles/second, or with a more common name, in

"Hertz" For example, the electric power we use in our daily life in the US is 60 Hz(50 Hz elsewhere in the world) This means that if you try to plot the electriccurrent, it will be a sine wave passing through the same point 50 times in 1 second.Now, look at the Figures below The first one is a sine wave at 3 Hz, the second one

at 10 Hz, and the third one at 50 Hz

Figure 1.19 Sine wave with frequencies at 3Hz, 10 Hz and 50 Hz [29]

In order to measure frequency, or find the frequency content of a signal, weused Fourier Transform (FT) If the FT of a signal in time domain is taken, thefrequency-amplitude representation of that signal is obtained In other words, wenow have a plot with one axis being the frequency and the other being the

amplitude This plot tells us how much of each frequency exists in our signal Thefrequency axis starts from zero, and goes up to infinity For every frequency, wehave an amplitude value For example, if we take the FT of the electric current that

we use in our houses, we will have one spike at 50 Hz, and nothing elsewhere, sincethat signal has only 50 Hz frequency component No other signal, however, has a

FT which is this simple For most practical purposes, signals contain more than onefrequency component The Figure 1.20 shows the FT of the 50 Hz signal:

Figure 1.20 The FT of 50Hz signal [29].

Often times, the information that cannot be readily seen in the time-domain can

be seen in the frequency domain Let us look at an ECG signal, which is an example

of biological signals The typical shape of a healthy ECG signal is well known tocardiologists Any significant deviation from that shape is usually considered to be asymptom of a pathological condition This pathological condition, however, maynot always be quite obvious in the original time-domain signal Cardiologists

usually use the time-domain ECG signals, which are recorded on strip-charts toanalyze ECG signals Recently, the new computerized ECG recorders/analyzersalso utilize the frequency information to decide whether a pathological conditionexists A pathological condition can sometimes be diagnosed more easily when thefrequency content of the signal is analyzed This, of course, is only one simpleexample why frequency content might be useful Today Fourier transforms are used

in many different areas including all branches of engineering

Although FT is probably the most popular transform being used (especially inelectrical engineering), it is not the only one There are many other transforms thatare used quite often by engineers and mathematicians Hilbert transform, short-timeFourier transform, Wigner distributions, the Radon Transform, and of course ourfeatured transformation , the wavelet transform, constitute only a small portion of ahuge list of transforms that are available at engineer's and mathematician's disposal.Every transformation technique has its own area of application, with advantages anddisadvantages, and the wavelet transform (WT) is no exception

FT (as well as WT) is a reversible transform, that is, it allows to go back andforwarding between the raw and processed (transformed) signals However, onlyeither of them is available at any given time That is, no frequency information isavailable in the time-domain signal, and no time information is available in theFourier transformed signal The natural question that comes to mind is that is itnecessary to have both the time and the frequency information at the same time As

we will see soon, the answer depends on the particular application and the nature ofthe signal in hand Recall that the FT gives the frequency information of the signal,which means that it tells us how much of each frequency exists in the signal, but itdoes not tell us when in time these frequency components exist This information isnot required when the signal is so-called stationary

The FT gives the spectral content of the signal, but it gives no informationregarding where in time those spectral components appear Therefore, FT is not asuitable technique for non-stationary signal However, if we are only interested inwhat spectral components exist in the signal, but not interested where these occur,

FT can be used for non-stationary signals In this case, if this information is needed,i.e., if we want to know, what spectral component occur at what time (interval) ,then Fourier transform is not the right transform to use For practical purposes, it isdifficult to make the separation, since there are many practical stationary signals, aswell as non-stationary ones Almost all biological signals, for example, are non-stationary Some of the most famous ones are ECG (electrical activity of the heart,electrocardiograph), EEG (electrical activity of the brain, electroencephalograph),and EMG (electrical activity of the muscles, electromyogram) Summarily, the FTgives what frequency components (spectral components) exist in the signal Whenthe time localization of the spectral components is needed, a transform giving theTime-frequency representation of the signal is needed.

The Wavelet transform is a transform of this type It provides the time-frequencyrepresentation (There are other transforms, which give this information too, such asshort time Fourier transform, Wigner distributions, etc Often times a particularspectral component, occurring at any instant can be of particular interest In thesecases, it may be very beneficial to know the time intervals these particular spectracomponents occur For example, in EEGs, the latency of an event-related potential

is of particular interest Event-related potential is the response of the brain to aspecific stimulus like flash-light, the latency of this response is the amount of timeelapsed between the onset of the stimulus and the response Wavelet transform iscapable of providing the time and frequency information simultaneously, hencegiving a time-frequency representation of the signal The WT was developed as analternative to the STFT The STFT will be explained in great detail in the secondpart of this tutorial It suffices at this time to say that the WT was developed toovercome some resolution related problems of the STFT To make a real long storyshort, we pass the time-domain signal from various high pass and low pass filters,which filters out either high frequency or low frequency portions of the signal Thisprocedure is repeated, every time some portion of the signal corresponding to somefrequencies being removed from the signal

1.3.2 Multi-resolution Analysis and Continuous Wavelet Transform

Although the time and frequency resolution problems are results of a physicalphenomenon and exist regardless of the transform used, it is possible to analyze anysignal by using an alternative approach called the multi-resolution analysis (MRA).MRA analyzes the signal at different frequencies with different resolutions Everyspectral component is not resolved equally as was the case in the STFT MRA isdesigned to give good time resolution and poor frequency resolution at highfrequencies and good frequency resolution and poor time resolution at lowfrequencies This approach makes sense especially when the signal at hand has highfrequency components for short durations and low frequency components for longdurations Fortunately, the signals that are encountered in practical application areoften of this type Figure 1.21 shows a signal of this type It has relatively lowfrequency component throughout the entire signal and relatively high frequencycomponents for a short duration somewhere around the middle

Figure 1.21 Example of a signal has high frequency components for short.

The Continuous Wavelet Transform

The continuous wavelet transform was developed as an alternative approach tothe short-time Fourier Transform to overcome the resolution problem The wavelet

analysis is done in a similar way to the STFT analysis, in the sense that the signal ismultiplied with a function, similar to the window function in the STFT, and thetransform is computed separately for different segments of the time-domain signal.However, there are two main differences between STFT and CWT:

- The Fourier Transforms of the windowed signals are not taken, and thereforesingle peak will be seen corresponding to a sinusoid, i.e., negative frequencies arenot computed

- The width of the window is changed as the transform is computed for everysingle spectral component, which is probably the most significant characteristic ofthe wavelet transform [29]

The continuous wavelet transform of a time series x (t) defined as follows:

ѱ (t) is the transforming function or mother wavelet with effective length (t) that

is usually much shorter than the target time series x (t) The term mother waveletgets its name due to two important properties of the wavelet analysis The term

“wavelet” means a small wave The smallest refers to the condition that thiswindow function is of finite length The wave refers to the condition that thisfunction is oscillatory The term mother implies that the functions with differentregions of support that are used in the transformation process are derived from onemain function, or the mother wavelet In other words, the mother wavelet is aprototype for generating the other window functions

The variables are s and τ, where s is the scale or dilation factor that determinesthe characteristic frequency so that its variation gives rise to a spectrum and τ is thetranslation in time so that its variation represents the sliding of the wavelet over x(t) The wavelet spectrum is thus customarily displayed in a time–frequencydomain For low scales, i.e., when |s|≪ 1, the wavelet function is highlyconcentrated (shrunken compressed) with frequency contents mostly in the higher

(1.3)