USF Libraries
USF Digital Collections

Arterial perfusion detection method by synchronous detection

MISSING IMAGE

Material Information

Title:
Arterial perfusion detection method by synchronous detection
Physical Description:
Book
Language:
English
Creator:
Prevot, Yohan
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Impedance plethysmography
Perfusion skin monitoring
Tissue impedance
Matlab
LabView
Dissertations, Academic -- Electrical Engineering -- Masters -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The pressure ulcer is a well-known clinical problem that has plagued many patients in acute-care hospitals and chronic-care facilities. The pressure ulcer has the potential to diminish the quality of a patient's life by hindering the person's physical and emotional well-being. In addition, pressure ulcers are a high-cost problem. Past studies have shown that costs related to the treatment of pressure ulcers have reached 1.335 billion dollars a year in the United States alone. A pressure ulcer is defined as a lesion created by unrelieved pressure, which causes tissue ischemia and subsequently damages the underlying tissue. This sequence of events is mainly centered on ischemia. Ischemia is a condition created by an insufficient flow of blood to an organ or part of an organ such as the skin. The outcome of ischemia is cell death at the tissue level, which is commonly termed necrosis. In the past, researchers employed several different non-invasive techniques in order t o detect changes in the condition of human skin when blood flow was restricted. The two most commonly used practices were Laser Doppler Velocimetry and Continuous Wave Ultrasound. Laser Doppler Velocimeter is used to measure cutaneous blood flow in a study region. The moving red blood cells in blood vessels cause a Doppler shift of incident laser light, which correlates with the velocity of blood flow. Continuous Wave Ultrasound involves an ultrasound signal, which is transmitted into the skin. The change in frequency of the reflected signal with respect to the transmitted signal provides an indication of blow flow. The objective of this research was to examine a method for the detection of arterial blood flow, which utilized the 4-electrode sensor for the measurement of Tissue Impedance or the Bio-impedance method. The system developed, for the synchronous detection method, consisted of both analog hardware and software tools. The analog hardware utilized synchronous detection. ^^The software monitored and performed mathematical operations on the retrieved data. The system developed during this research demonstrated the ability to measure the pulsatile impedance of the skin and present the results in a fashion useful to healthcare providers.
Thesis:
Thesis (M.S.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Yohan Prevot.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 92 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001912823
oclc - 174113287
usfldc doi - E14-SFE0001385
usfldc handle - e14.1385
System ID:
SFS0025705:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Arterial Perfusion Detection Method By Synchronous Detection by Yohan Prevot A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Electrical Engineering Department of Electrical Engineering College of Engineering University of South Florida Major Professor: Wilfrido Moreno, Ph.D. Co-Major Professor: Jeffrey Harrow, Ph.D. James T. Leffew, Ph.D. Paris Wiley, Ph.D. Horace Gordon, M.S.E.E. Date of Approval November 4th, 2005 Keywords: Impedance Plethysmography, Perfus ion Skin Monitoring, Tissue Impedance, Matlab, LabView Copyright 2005, Yohan Prevot

PAGE 2

DEDICATIONS To God To Judy Etzel; for keeping me grounded throughout the years To Pam and Kefton Schermerhorn; for alwa ys keeping me optimistic and patient To Ryan and Afra Beggy; for bei ng there at the time most needed To Lauren Byrd; the birth of you changed my life To Eduardo Zurek; for helping me throughout my engineering years To Nhat Nguyen; for finishing our bachelors and masters together To Bill Rothembach; for starting engineering school with me

PAGE 3

ACKNOWLEDGEMENTS I extend my gratitude to Dr. Moreno and Dr Leffew for being my professors and mentors. In addition, I would like to thank them for providing me with guidance and the motivation to excel in engin eering, throughout my studies at the University of South Florida. I sincerely appreciate and thank Dr. W iley and Prof. Gordon for always making time to help me with my designs. Dr. Harrow provided me the opportunity to work on the Impedance project. He was always concerned, supportive, helpfu l and patient. Thank you Dr. Harrow. I must also recognize Eduado Zurek and Nhat Nguyen for being my fellow engineers and most of al l my friends. Thank you!

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES iv LIST OF FIGURES v ABSTRACT ix CHAPTER 1 INTRODUCTION 1 1.1 Tissue Ischemia and Pressure Ulcers 1 1.2 Precedent Studies on Arterial Perfusion Detection Method 4 1.2.1 Impedance Plethysmography 4 1.2.1.1 Tissue Impedance 5 1.2.2 Photo-Plethysmography 7 1.3 Laser Doppler Velocimetry 8 1.4 Motivation and Thesis Framework 9 CHAPTER 2 BACKGROUND THEORY 11 2.1 Arterial Perfusion 11 2.2 Skin Anatomy and Physiology 11 2.2.1 Electrical Properties of Tissue 13 2.3 Ag/AgCl Electrodes 14 2.4 The Four-Electrode Measurement System 15 CHAPTER 3 ANALOG HARDWARE DESCRIPTION 17 3.1 Analog Hardware 17 3.1.1 Function Generator 18 3.1.2 Programmable Power Supply 18 3.1.3 Second Order, Narrow-Band, Bandpass Filter Based On Inductor Replacement 20 3.1.3.1 Circuit Analysis 21 3.1.3.2 PSPICE Software Simulation 22 3.1.3.3 Implementation Results 24 3.1.4 Constant Current Source 26

PAGE 5

ii 3.1.4.1 Circuit Analysis 27 3.1.4.2 PSPICE Software Simulation 30 3.1.4.3 Implementation Results 32 3.1.5 Emitter Follower 32 3.1.5.1 Circuit Analysis 33 3.1.5.2 PSPICE Software Simulation 34 3.1.6 BJT Differential Amplifier and Buffer 36 3.1.6.1 Circuit Analysis 37 3.1.6.2 PSPICE Software Simulation 38 3.1.7 Second Order Phase Shifter with Voltage Divider and Op-Amp Buffer 39 3.1.7.1 Circuit Analysis 40 3.1.7.2 PSPICE Software Simulation 41 3.1.8 AD630 Integrated Circuit 43 3.1.8.1 Synchronous Detection 44 3.1.8.2 PSPICE Software Simulation 45 3.1.8.3 Implementation 47 3.1.9 Sallen & Key Second Order Active Highpass Butterworth Filter 49 3.1.9.1 Circuit Analysis 51 3.1.9.2 PSPICE Software Simulation 51 3.1.10 AD620 Integrated Circuit 53 3.1.10.1 Circuit Analysis 54 3.1.10.2 PSPICE Software Simulation 55 3.1.11 Third Order Passive RC Filter 56 3.1.11.1 Circuit Analysis 57 3.1.11.2 PSPICE Software Simulation 57 CHAPTER 4 SOFTWARE DESCRIPTION 59 4.1 Software Description 59 4.1.1 National Instruments LabView 59 4.1.2 Data Acquisition 60 4.1.3 LabView Interface 61 CHAPTER 5 TESTING AND RESULTS 63 5.1 Forearm Limb Simulation 63 5.1.1 Arm-Cuff: Deflated 64 5.1.2 Arm-Cuff: Inflated and Released 66 5.2 Ankle Limb Experimentation 69 5.2.1 Arm-Cuff: Deflated 70 5.2.2 Arm-Cuff: Inflated and Released 72 5.2.3 Motion Artifacts 76

PAGE 6

iii CHAPTER 6 CONCLUSION AND RECOMMENDATIONS 79 6.1 Conclusions 79 6.2 Recommendations 80 REFERENCES 81 APPENDICES 83 Appendix A Matlab Source Code for the Transfer Function 84 Appendix B Matlab Source Code for the Arterial Perfusion Detection Method 86 Appendix C Constant Current Source DC Currents and Node Voltages 88 Appendix D Trouble Shooting AC Volta ge to AC Current Converter 89 Appendix E Design of the Emitter Follower 91

PAGE 7

iv LIST OF TABLES Table 1.1: Resistivity Values for Se lected Human Body Components 6 Table 3.1: Calculated and Simulated 3dB Bandwidth and Quality Factor 23 Table 3.2: DC Characteristics fo r the BJT Simulation 30 Table 3.3: Transistor Operati ng Currents and Voltages 35

PAGE 8

v LIST OF FIGURES Figure 1.1: Area Where Bedsores Occur 2 Figure 1.2: Transverse and Longitudinal Impe dances of the Skeletal Muscle 7 Figure 1.3: Schematic Illustration of the PPG Function 8 Figure 1.4: Laser Doppler Mechanism 9 Figure 2.1: Layers of Human Skin 13 Figure 2.2: High and Low Frequency Current Paths in Tissue 14 Figure 2.3: Four Electrode Technique 15 Figure 3.1: Block Diagram for the Ar terial Perfusion Detection Method 17 Figure 3.2: PS2521G Programmable Power Supply 19 Figure 3.3: Second Order, Narrow-Ba nd, Bandpass Filter with Inductor Replacement 20 Figure 3.4: Simulated Frequency Response of the Filter 22 Figure 3.5: Pole/Zero Plots for the Sec ond Order, Narrow-Band, Bandpass Filter 24 Figure 3.6: Frequency Response of th e Implemented Bandpass Filter 25 Figure 3.7: Analysis of the Quality Factor 26 Figure 3.8: Schematic Diagram of th e Constant Current Source 27 Figure 3.9: Top Half of the Cons tant Current source 28 Figure 3.10: Analysis of the Top Half of the Constant Current Source 29

PAGE 9

vi Figure 3.11: Bottom Half of the Current Source 30 Figure 3.12: Output Voltage with Fluctuating Load Resistor 31 Figure 3.13: Output Current with Fluctuating Load Resistor 31 Figure 3.14: Measured Data from the Multi-Meter 32 Figure 3.15: Emitter Follower 33 Figure 3.16: DC Node Voltages and Currents 34 Figure 3.17: Frequency Response of the Emitter Follower 35 Figure 3.18: Schematic Diagram of the BJ T Differential Amplifier and Buffer 36 Figure 3.19: Gain and CMRR Analysis of the Differential Amplifier 37 Figure 3.20: DC Currents of the Simulated Differential Amplifier 38 Figure 3.21: Gain Response of the Differential Amplifier 39 Figure 3.22: Schematic Diagram of the Second Order Allpass Filter Voltage Divider and Buffer 40 Figure 3.23: Calculation of the Compone nt Values for a Phase Shifter 41 Figure 3.24: Second Order Allpass Phas e Shifter Simulation Results 42 Figure 3.25: Verification of the Phase-Shifter Design 42 Figure 3.26: Allpass Phase Shifter Pole/Zero Plots 43 Figure 3.27: Functional Block Diagram of the AD630 44 Figure 3.28: AD630 as a Gain-Of-One Balanced Modulator 45 Figure 3.29: PSPICE Schematic of the Simulation 46 Figure 3.30: PSPICE Generated, 100 kHz Square Wave 46 Figure 3.31: PSPICE Generated, 10 kHz Sine Wave 47

PAGE 10

vii Figure 3.32: AD630 Output Simulation 47 Figure 3.33: 100 kHz Square Wave 48 Figure 3.34: 10 kHz Sine Wave 48 Figure 3.35: Implementation Output from the AD630 48 Figure 3.36: Data Sheet Results 49 Figure 3.37: AD630 Simulation Output 49 Figure 3.38: AD630 Implementation Output 49 Figure 3.39: Sallen & Key Highpass Filter 50 Figure 3.40: Analysis of the Highpass Filter 51 Figure 3.41: Simulated Frequency Re sponse of the Highpass Filter 52 Figure 3.42: Pole/Zero and Gain Re sponse of the Highpass Filter 53 Figure 3.43: Functional Block Diagram of the AD620 54 Figure 3.44: Analysis of the Gain Resistor 55 Figure 3.45: PSPICE Schematic of the AD620 55 Figure 3.46: AD620 IC Simulation Output 56 Figure 3.47: Third Order Passive Lowpass RC Filter 56 Figure 3.48: Analysis of the RC Filter Components 57 Figure 3.49: Passive Lowpass Filter Frequency Response 58 Figure 4.1: Illustration of the La bView Programming Style 60 Figure 4.2: Data Acquisition System 60 Figure 4.3: Software Interface for the Arte rial Perfusion Detection Circuit 61 Figure 4.4: LabView Graphical Codes 62 Figure 5.1: Test Electrode Placements 64

PAGE 11

viii Figure 5.2: Setup for the Forearm Limb Measurement 64 Figure 5.3: Arterial Pulse from Arterial Perfusion and Photo-plethysmograph 65 Figure 5.4: An Arm-Cuff 66 Figure 5.5: Simulation with an Arm-Cuff 66 Figure 5.6: Original an d Derivative Waveforms of the APD Design 68 Figure 5.7: Anatomy of the Forearm 69 Figure 5.8: Ankle Limb Experimentation 70 Figure 5.9: APD and Ph oto-Plethysmograph Signals 71 Figure 5.10: Simulation with the Arm Cuff 72 Figure 5.11: Initial Pulse of the Ankle Experiment 73 Figure 5.12: Pulsatile Impedance with the Arm Cuff Fully Inflated 74 Figure 5.13: Original Filtered APD Si gnal and the Signals’ Derivative 74 Figure 5.14: Initial Pulsatile Cha nges and Their Derivative 75 Figure 5.15: Thorax Impedance Curve 75 Figure 5.16: Output Signal of the Arteri al Perfusion Circuit with Movement 76 Figure 5.17: Effects of the Initial Movement of the Big Toe 77 Figure C.1: BJT Node Curre nts and Voltages 88 Figure D.1: Constant Current Source 89 Figure E.1: Emitter Follower Schematic 91 Figure E.2: Small Signal Analysis 91

PAGE 12

ix ARTERIAL PERFUSION DETECTION METHOD BY SYNCHRONOUS DETECTION Yohan P. Prevot ABSTRACT The pressure ulcer is a well-known clin ical problem that has plagued many patients in acute-care hospitals and chronic-care facilities. The pressure ulcer has the potential to diminish the quali ty of a patient’s life by hinderi ng the person’s physical and emotional well-being. In additi on, pressure ulcers are a high-co st problem. Past studies have shown that costs related to the treat ment of pressure ulcers have reached 1.335 billion dollars a year in the United States alone. A pressure ulcer is define d as a lesion created by unrelieved pressure, which causes tissue ischemia and subsequently dama ges the underlying tissue. This sequence of events is mainly centered on ischemia. Ischemia is a condition created by an insufficient flow of blood to an organ or part of an organ such as the skin. The outcome of ischemia is cell death at the tissue level, which is commonly termed necrosis. In the past, researchers employed severa l different non-invasive techniques in order to detect changes in the condition of human skin when blood flow was restricted. The two most commonly used practices were Laser Doppler Velocimetry and Continuous Wave Ultrasound. Laser Doppler Velocimeter is used to measure cutaneous blood flow

PAGE 13

x in a study region. The moving red blood cells in blood vessels cause a Doppler shift of incident laser light, which corre lates with the velocity of blood flow. Continuous Wave Ultrasound involves an ultrasound si gnal, which is transmitted in to the skin. The change in frequency of the reflected signal with respect to the transmitted signal provides an indication of blow flow. The objective of this research was to examine a method for the detection of arterial blood flow, which util ized the 4-electrode sensor for the measurement of Tissue Impedance or the Bio-impedance method. Th e system developed, for the synchronous detection method, consisted of both analog ha rdware and software tools. The analog hardware utilized synchronous detection. The software monitored and performed mathematical operations on th e retrieved data. The syst em developed during this research demonstrated the ability to measur e the pulsatile impedance of the skin and present the results in a fashion useful to healthcare providers.

PAGE 14

1 CHAPTER 1 INTRODUCTION 1.1 Tissue Ischemia and Pressure Ulcers The purpose of this chapter is to introduce the basic mechanisms related to the formation of pressure ulcers and the tools that have been designed, to date, for the detection of blood flow. In the developed world today one of the major causes of morbidity and mortality is ischemia. Ischemia is a condition that arises when tissue substrates requirements for energy, O2 and glucose, are not met by supply due to inadequate arterial perfusion. This circumstance has led health organizations to combine their efforts to better educate clinicians and, with the assist ence of engineers, to devise medical instrumentation for the detection of blood flow. Ischemia of the skin occurs when external pressure compresses the blood vessels. Compression of the blood vessels causes an occlusion of blood flow. Prolonged occlusion leads to the devel opment of pressure ulcers, which are most commonly found in patients with spinal cord injury. Pressu re ulcers usually develop in soft tissues over bony prominences that remain in contact with an applied pressure to the surface. Such stress on the soft tissue eventually reduces blood supply to the sk in and the clinical

PAGE 15

2 symptom of localized tissue death, (necrosis), results. Pressure affects many different body tissues since they exhibit different tolera nces for pressure, acti ng force on a specific area and ischemia. For example, muscle tissue e xhibits a greater sensitivity to stress than the skin organ. Unrelieved pressure applied to the skin surface ex erts its greatest force near the bone. Therefore, se rious damage may o ccur between skin and bone before it becomes apparent that and when the skin is broken, [22]. Additional factors other than pressure such as sheer, infecti on and friction also play a role in the mechanical damage of the skin. The variations in the formation of pressure ulceration have ultimately been structured to better evaluate the timeline for their existence so that timely and adequate care can be provided. Assessment of a pressure ulcer is classified in stages. The initial stage, (stage 1), involves the observation of reddened ski n, which does not blanch. Stage 2, is characterized by the appearance of a bliste r. Stage 3 is entered when the blister deteriorates into open wound, which presents ne crosis of the subcutaneous tissue. Stage 4 exists when tissue necrosis down to the bone is presented, [14]. Figure 1.1 illustrates one of the most common locations wh ere a pressure ulcer might develop. Figure 1.1: Area Where Bedsores Occur, [15]

PAGE 16

3 The treatment and prevention of pressure ulcerations is a tedious task. Many guidelines such as pressure management were created for clinical personnel to better assist in the protection of th e patient skin and deep tissues Pressure management is related to body positioning, which is defined by three categories: 1. In Bed, 2. Turning and repositioning, 3. Sitting. When lying in bed, pillows should be placed between the legs so that the knees and ankles do not exert pressure on each other. Turning and repositioning is extremely important for spinal cord injury patient. Pa tients with spinal co rd injury lack active sensory nerves for the detect ion of local pressure, while a healthy individual will normally move whether awake or sleeping. The third condition of concern related to pressure ulceration occurs w ith the sitting patient sin ce sitting involves increased pressures on smaller sacral t uberosity region, which produces a greater risk of pressure ulcer formation. Therefore, an appropriate body posture and alignment of the limbs must be maintained. Additional instructions include skin care, wound management and nutrition, which can be obtained from, [22]. Pressure ulcers are not a simple problem with an easy solution. Continuous care and monitoring of the patient, whether clinical or emotional, must be provided in order to insure successful healing of the tissue a nd the prevention of any re-occurrences.

PAGE 17

4 1.2 Precedent Studies on Arterial Perfusion Detection Applications Over decades, medical doctors and Bio-Instrumentation engineers have collaborated in the design of medical equipm ent to detect blood flow in the human body, which could warn of ischemia and preven t the ischemia from causing irreversible necrosis, thus promote a nursing or patient-cen tered intervention, such as weight-shifting. The following subsection illustrates some of the current applications that were studied in order to devise a new concept for the detection of Arterial Perfusion. 1.2.1 Impedance Plethysmography Plethysmography is defined as the measurement of volume. This term is often associated with volume changes due to the bea ting of the heart but may also be related, for example, to respiration or peristalti c movements of the alimentary canal [8]. Measurements of these changes can be done by the use of strain gauges, impedance and various other methods. Impedance plethysmography is a method of determining changing organ volumes in the body, based on the measurement of electric impedance at the body surface It has been developed and commercially marketed for measuring cardiac output. The volume change detected is the difference in heart volume between systole (full) and diastole (empty). Measurement of transthoracic impe dance can thus monitor the changes due to the cardiac cycle. It can also be used to detect outflow obstruction in the veins of the lower extremities. By putting a tourniquet ar ound a leg and inflating it to a pressure higher than venous pressure, but below arteri al, blood will flow into the extremity, filling it up like a balloon. When the tourniquet is re leased, the blood flows out, and the rate of

PAGE 18

5 that flow can be measured by watching the change of impedance of the extremity, [8]. For the purpose of this paper, Tissue Impe dance, the physical quantity measured in impedance plethysmography, will be furthered examined for the detection of arterial perfusion in extremities. 1.2.1.1 Tissue Impedance The human body may be considered as a resistive, piecewise homogeneous and linear volume conductor [9]. Organs, muscle tissue, nerve tissues all exhibits unique resistive and conductive propert ies. Below, Table 1.1 illu strates the various tissue resistivity values, physical quantity measured in impedance plethysmography, of a number of components of the human body. The values shown in table 1.1 can be used to calculate the impedance of the blood volume with blood resistivity b based on the geometrical diameter of the lim bs using the formula, z = ( *L)/ A where is the blood resistivity, A is the cross sect ion of the area of blood and L is the length of the limb in study [9].

PAGE 19

6 Table 1.1: Resistivity Values for Selected Human Body Components, [9] Tissue [ m] Remarks Reference Brain Cerebrospinal fluid Blood Plasma Heart muscle Skeletal muscle Liver Lung Fat Bone 2.2 6.8 5.8 0.7 1.6 0.7 2.5 5.6 1.9 13.2 7 11.2 21.7 25 177 15 158 215 gray matter white matter average Hemaatocrit = 45 longitudinal transverse longitudinal transverse longitudinal circumferential radial (at 100 kHz) Rush and Driscoll, 1969 Barber and Brown, 1984 Barber and Brown, 1984 Geddes and Sadler, 1973 Barber and Brown, 1984 Rush, Abildskov, and McFee, 1963 Epstein and Foster, 1982 Rush, Abildskov, and McFee, 1963 Schwan and Kay, 1956 Rush, Abildskov, and McFee, 1963 Geddes and Baker, 1967 Rush and Driscoll, 1969 Saha and Williams, 1992 Tissue impedance can also demonstrate frequency dependence since the impedance of different tissues exhibit reactiv e properties under cert ain conditions. This dependence must be considered when selec ting the operating frequency for the design a Bio-Impedance system. One method that is us ed to demonstrate the properties of tissue impedance as a function of frequency is th e Cole-Cole plot. Figure 1.2 presents an example of the Cole-Cole plot of the tissue impedance behavior of skeletal muscle as a function of frequency.

PAGE 20

7 Figure 1.2: Transverse and Longitudinal Impedances of Skeletal Muscle, [9] The Cole-Cole plot has proven to be useful for the determination of tissue impedance corresponding to different parts of the body. Further analysis of this method can be found in the literature, [9]. 1.2.2 Photo-Plethysmography A Photo-Plethysmograph, (PPG), is a devi ce, constructed from a light source and a detector, for the purpose of measuring bl ood volume in various parts of the body. The PPG captures the pulsation of the arterial bl ood flow, which is the AC component of the signal retrieved from the photo-detector. The sites around the body that are commonly used for detecting such volume changes include the finger, the ear lobe and the foot. There are two different method employed wh en using a photo-plethysmograph, one can be transmissive or reflective. In simple PPG the absorption, reflection and scattering of the induced light is dictated by the volume of blood in the vessels due to arterial pulsation. Measurement of the varying light that is either transmitted or reflected will show changes in the amount of blood in the body part. A more complic ated system, pulse oxymetry, looks

PAGE 21

8 simultaneously at two different wavelengt hs, which are based on the spectroscopic differences between oxyhemoglobin and de oxyhemoglobin. The use of two different wavelengths provides for the determination of arterial oxygen satura tion and pulse rate. Figure 1.3: Schematic Illust ration of the PPG Function 1.3 Laser Doppler Velocimetry Laser Doppler Velocimetry provides a noninvasive capability for the continuous measurement of capillary blood flow. This method has been used to measure capillary blood flow in many tissues including muscle, skin, bone and intestine. Laser Doppler Velocimetry entails the use of, minimal power, laser light that is transmitted via an optic fiber placed on top of the area of the skin under consideration. Upon contact with the skin, the light is sc attered by moving red blood cells, which changes the frequency of the light according to the Doppler Effect. The scattered light is transmitted through optical fibers from the subject to an analyzer that displays the Laser Doppler flux signal. The output of the analyzer presents the value for the flux of red cells, which is the number of red cells times their velo city. Further literature related to this technique can be found in the liturature, [16]. Figure 4 illustrates the basic mechanism employed by Laser Doppler velocimetry. Light Source Bed Tissues, Finger Detector

PAGE 22

9 Figure 1.4: Laser Doppl er Mechanism, [16] 1.4 Motivations and Thesis Framework The motivation for this research was the need to examine alternative methods for the detection of arterial blood flow for the purpose of preven tion of pressure ulcers in spinal cord injury patients. Pressure ulcer s result in high care co st and high rates of reoccurrence since no methods, specifically designed for pr essure ulcers prevention by monitoring skin blood flow, exit. Many prove n concepts for the measurement of skin blood flow, currently being used, cannot be used for the purpose of prevention of pressure ulcers in patients. Therefore, it was apparent that a need existed for a small, cost effective device, which could be attached to a patient in or der to continuously monitor blood circulation.

PAGE 23

10 This thesis will discuss the anatomy of th e skin and its electrical properties in Chapter 2. An understanding of how the skin behaves under electrical stimulation will disclose why the design concept, used to complete the Arterial Perfusion Method, was chosen. In Chapter 3, the analog theory, so ftware simulation and system implementation are discussed. Chapter 4 presents the software interface that was designed to monitor, in real time, the data obtained. Chapter 5 presents subject data in order to demonstrate how the device performed when placed on vari ous parts of the body. Conclusions and recommendations for future investigations are presented in Chapter 6.

PAGE 24

11 CHAPTER 2 BACKGROUND THEORY 2.1 Arterial Perfusion In physiology, arterial perf usion is the process concer ned with the delivery of nutrients by means of blood flow to a capillary bed in the biological tissue or the limb. The degree to which blood is delivered to tissues plays an important ro le in the healing of pressure ulcers, which are prevalen t in patients confined to beds. 2.2 Skin Anatomy and Physiology The human skin is the largest organ of the human body and provides protection for the body from the outside world. Any damages or impairment of the skin has important implications for the health of a pe rson. Its function is to regulate the body temperature, transmit the sensation of touch and pain, prevent loss of bodily fluids and protect the inside of the human body against incursion. Essentially, the skin is composed of tw o layers, which are termed the epidermis and the dermis. The epidermis is the outmost layer of the skin, which is continually renovating itself through the flaking of aged cells and attaining new cells from the dermis

PAGE 25

12 region. The dermis is the layer where sweat glands, hair cells a nd nerve receptors are found. The stratum corneum, (SC), which is the s ubcutaneous layer of the epidermis, was one of the main areas for elec trical consideration. An unde rstanding of the properties of the SC, under electrical stimulation, was of great importance for the successful implementation of the arterial perfusion detection method. The stratum corneum is the outermost laye r of the epidermis. Its surface is composed of dead cells while healthy livi ng cells are found at its base, [9]. The physiological function of the stratum corneu m is to provide the body with a defense against mechanical and chemical attacks from the outside world. A dditionally, its water insulating fatty composition also acts as a hi gh resistance to electrical signals with DC components. Apart from its protection from real world conditions, the stratum corneum provides a medium for the study of biological signals. The human body contains many electrical si gnals that can be captured by the placement of skin electrodes on the stratum corneum. These signals originate from nerve impulses or muscle contraction. The SC of the skin presents a very high electrical impedance. Therefore, high input-impedance amp lifiers must be used for the detection of such signals. The skin and its vari ous layers are depicted in Figure 2.1.

PAGE 26

13 Figure 2.1: Layers of the Human Skin, [17] The stratum corneum provide s the dominant skin impedance at frequencies less than ten kilo-Hertz. Its co rresponding impedance was the determining factor in the decision to use a frequency of 100 kHz for the arterial perfusion dete ction circuit. In turn, the use of a signal with a 100 kHz fre quency component for the arterial perfusion detection circuit led to a deeper analysis of the layers beneath the stratum corneum as well as the reduction in the impe dance of the stratum corneum. 2.2.1 Electrical Properties of Tissue Tissue, a body component made of livi ng cells, exhibits frequency dependent electrical behavior that is determined by the conductivity of the intra-cellular spaces, extra-cellular spaces and the permittivity of th e membrane, [12]. Electrical current flow in the body tissue is a function of the state of the tissue, the ti ssue structure and the current source employed. Current flow in healthy tissue follows the principle: At dc and low frequencies, current applied to the body passes around the cell since the cell membrane has high capacitive properties. With frequencies larger

PAGE 27

14 than one kilo-hertz the current appl ied will pass through the cells since the capacitive property of the cell membra ne no longer affects its path. An example of the current pattern in vi able tissue is demonstrated in Figure 2.2. Figure 2.2: High and Low Freque ncy Current Paths in Tissue In general, tissue in the body, whether musc le tissue or nerve tissue, will behave differently to certain frequencies. For exampl e, the longitude direction in muscle tissue possesses high conductance and is not freque ncy dependent. These characteristics indicate that direct free liquid chan nels dominate the current path, [8]. 2.3 Ag/AgCl Electrodes The silver-silver chloride electrode is the most common electrode due to its straightforwardness of manufacture and its temp erature range. Silv er-silver chloride, (Ag-AgCl), electrodes provide precise transm ission of surface bio-potentials developed across the human skin. Only pure silver-sil ver chloride is used in measurements associated with human skin. During fabricat ion a bendable silver wi re is incorporated into the Ag-AgCl matrix to pr ovide an electrical link. An adhesive, (wet gel), is also added to allow for good contact between the skin and the electrode. HF LF

PAGE 28

15 2.4 The Four-Electrode Measurement System A diagram of the four-electrode system for the measurement of tissue impedance is presented in Figure 2.3. Iin V+ VI out Skin Layer Equi-potential surface Figure 2.3: Four-Electrode Technique The four-electrode measurement system has been widely used by researchers since it provides certain benef its with respect to its counter-part, which involves the use of just two electrodes. A major differe nce between the two-electrode measurement method and the four-electrode measurement method is: 2-electrode-measurement: Causes errors since the potential difference sensed between the two electrodes includes the voltage due to the current flowing through the polarizat ion impedance at the electrodetissue interface, [13]. 4-electrode-measurement: An AC si gnal is passed through the outer two electrodes and the inner two electrodes sense the voltage difference.

PAGE 29

16 Therefore, utilization of four-electrodes introduces: 1) Minimal error due to electrode-tissue interface impedance, 2) Effects of polarization impedance are negligible when using high input impedance amplifiers. When building a device to u tilize the four-electrode me thod two key issues related to the injected current and electrode spacing must be addressed. The amplitude of the injected current must be determined since its amplitude will give rise to a decrease in the impedance of the SC and will lead to the analys is of deeper layers. Another factor is the spacing of the electrodes. Electrode placemen t needs to be linear and equally spaced on the subject. Overall, the four-electrode m easurement is best suited for the measurement of tissue impedance and provides the best pos sible sensor for the arterial perfusion detection method.

PAGE 30

17 CHAPTER 3 ANALOG HARDWARE DESCRIPTION 3.1 Analog Hardware This chapter contains a description of the analog hardware used in this research. Additionally, this chapter explains the theor y, software analysis and implementation of the various components used in order to real ize the Arterial Perfusion Detection method. Figure 3.1 presents the overall block diagram of the final implementation of the arterial perfusion detection design Narrow Band-Pass Filter Fc = 100 kHz Function Generator 100 kHz, 5V Constant Current Source 100 kHz, 500uA Differential Amplifier Gain = 5 V/V Buffers Buffer Phase Shifter 180o Synchronous Detector Filter High Pass Fc = 500 mHz Filter Low-Pass Fc = 3Hz Amplifier Gain = 1000 4-Electrode System I+ V+ VIOutput Figure 3.1: Block Diagram for th e Arterial Perfusion Detection Method

PAGE 31

18 3.1.1 Function Generator The AFG300 instrument is a 16 MHz arbi trary waveform generator with built-in waveform capabilities. The instrument suppor ts standard functions such as sine wave, square wave, triangle wave, DC levels and noise Additional features incorporated in this instrument are: Function Generator, Burst Generator, Sweep Generator, Modulation Source. Applications for this instrument exist for such diverse areas as the classroom laboratory and simulating waveforms in testi ng and design for biomedical purposes. For this research, the function gene rator was used to generate an input sine wave, with a frequency of 100 kHz, for the Arteri al Perfusion Detection circuit. 3.1.2 Programmable Power Supply The PS2521G is a triple-output Power Supply with one supply delivering a limit of 6V and 3A and two supplies delivering a maximum of 36V and 1.5A. The power supplies can be controlled inde pendently. In the independe nt mode, the output voltage and current of each supply are controlled indi vidually and independent of the other two supplies. In the two tracking modes the variab le outputs are connected either in series or in parallel and the controls of the master pow er supply adjusts the voltage or current of both power supplies, [18]. This power s upply can be used to provide power to electronics components when designing, perfor ming test during laboratory research or for

PAGE 32

19 education related objectives. The PS2520G Programmable Power Supply was used in this research to provide positive and negative 12 Volt power to the Arterial Perfusion Detection circuit. Figure 3.2 pictures th e front panel of the PS2521G power supply. Figure 3.2: PS2521G Programmable Power Supply, [18]

PAGE 33

20 3.1.3 Second Order, Narrow-Band, Bandpass Filter Based On Inductor Replacement For years, analog design engineers have proposed ways to replace inductors with op-amps in electronic circuits. One of the best realizations came from A. Antoniou, (1969). His design has proven to be very tolerant to the non-ideal properties of the opamps included in the circuit. In particular the non-ideal properties associated with the opamps finite gain and bandwidth, [4]. The decision to use the op-amp based filters instead of inductor based filters was, essentially, personal. The fact is that problem s can arise, with the use of inductors, as a result of their lossyness, which is due to significant series re sistance and many other pathologies. Figure 3.3 presents the Ant oniou inductance model circuit for a Second order, narrow-band, bandpass filter. G V-R12 1.5k G G U2 AD817/AD 3 2 7 4 6 + -V+ V-OUT V-C10 .1u R14 2000 G V++ U3 AD817/AD 3 2 7 4 6 + -V+V-OUT R13 1.5k R15 250 Rvariable1 700 G U1 AD817/AD 3 2 7 4 6 + -V+V-OUT V++ C6 .1u C9 .1u C7 1000p R16 100 C5 .1u C8 .1u C3 1000p G C4 .1u V-G R10 332k G G R11 1.5k V++ Rvariable 1.49k Vin Figure 3.3: Second Order, Narrow-Band, Ba ndpass Filter with Inductor Replacement For this research, the bandpass filter was utilized to f ilter out any harmonics that were generated by the function generator and 60Hz power line frequency. To achieve a

PAGE 34

21 condition as close as possible to a system without large ma gnitude harmonics, a quality factor, (Q), greater than twen ty was desired for the filter. 3.1.3.1 Circuit Analysis The second order, narrow-band, bandpass filt er had to satisfy requirements of: Center Frequency of 100 kHz, Q > 20, Center Frequency Gain of 1, C = 1000pF, 12 Volt supply. The center frequency chosen for the design of the bandpass filter had to coincide with the implementation of a constant cu rrent source operating at that frequency. Additionally, a high Q factor, gr eater than 20, was desired so that the narrow-band filter frequency response could be obtained. The computation of the resistor values and bandwidth values were performed by using equation (1), (2), (3 ) and the arbitrarily chosen capacitor value. Pole Frequency Equation R 1 R 2 R 3 R 5 1 o C (1) Pole Q Factor Equation R 6 Q o C (2)

PAGE 35

22 3-dB Bandwidth Equation Bf h f l (3) 3.1.3.2 PSPICE Software Simulation After all the required component values we re calculated, the design was simulated in Orcad PSPICE. The PSPICE design software tool is the most widely used schematic entry system presently employed in electroni c design. It specifically allows for the simulation of analog circuits and the revision of a printed circuit board, (PCB), diagram. The main objective for simulating the filter circuit was to observe its frequency response so that the 3-dB bandwidth and the quality f actor could be determined and compared to the calculated value. Figure 3.4 graphically presents the simulation results for the filter’s frequency response. Frequency 100KHz 80KHz 120KHz V(Rvariable1:2) 0V 0.5V 1.0V Figure 3.4: Simulated Freque ncy Response of the Filter

PAGE 36

23 Using equation (2) and (3) the 3-dB bandw idth and the Quality Factor of the simulated frequency response was determine d. The results of the calculations are presented in Table 3.1. Table 3.1: Calculated and Simulated 3-dB Bandwidth and Quality Factor Calculated Simulated 3-dB Bandwidth 3.012e3 3.677e3 Quality Factor (Q) 208.602 27.225 The Quality factor from the simulation varied significantly from the theoretical calculation. This result was to be expected since the theoretical aspect of the design did not take into consideration the parasitic eff ect of the capacitors. Another aspect of the theoretical design of the filter, which needed to be evaluated, was the filter’s pole/zero placements, as well as its Gain response. In order to investigate and establish these aspects of the design Matlab software was employed. Matlab is a data-manipulation tool that allows data synthesis by user-designed programs. Matlab software was employed to study the characteristics of th e filter. The second order, narrow-band, bandpass filter has a second order transfer f unction that is given by: Ts ()Ks C 6 R 6 s2s 1 C 6 R 6 R 2 C 4 C 6 R 1 R 3 R 5 (4) Analyses of the transfer function with th e Matlab source code, which is presented in Appendix 1, generated the data presented in Figure 3.5. The top graph presents the pole/zero plot, which shows th at the poles were located 0.1500 rad from the imaginary axis. Using the distance of the poles from the jw-axis, the quality factor, (Q), was

PAGE 37

24 determined to have a negative value. A ne gative Q value implied that the poles were located in the right half of the s-plan e which would produce oscillations, [4]. Figure 3.5: Pole-Zero Plots of the Second Order, Narrow-Band, Bandpass Filter The bottom graph in Figure 3.5 presents the filter’s Gain response. High attenuation, in the orders of -100dB, can be achieved outside of the 3dB bandwidth. Overall the filter design met the requirements specified for the Arterial Perfusion Detection Method. 3.1.3.3 Implementation Results The final step of the design, after havi ng calculated its frequency response and pole/zero placements, was to construct the filte r. Testing of the physical implementation of the filter demonstrated, through its overall frequency response, that the filter could actually be constructed for a real world appli cation. Using a function generator, an input amplitude of 1V for frequencies in the range of 90 kHz to 110 kHz was supplied to the filter. The output amplitude was recorded and the gain of the filter calculated and plotted. Figure 3.6 presents the re sults from four different tests.

PAGE 38

25 9 1049.5 1041 1051.05 1051.1 1050 0.5 1 1.001 0.054 gain data1 gain 2 data m1 1.1105 9104 frequency Figure 3.6: Frequency Response of Implemented Bandpass Filter The variation in the re sponses at different frequencies arose due to implementation of the circuit with tantalum and mica capacitors. The test, using different kinds of capacitors, was performed to obser ve whether the frequency response would be affected since each type of capacitor exhibits different fr equency response properties. Tantalum capacitors have less series resist ance, higher capacitance to volume ratio and are stable. Mica capacitors have low series resistance, inductance and are generally used for high frequency filtering. In Figure 3.6, Data<1> was the original data whose amplitude was recorded in terms of its root mean square, (RMS), value while the gain was normalized. Gain2 was also normalized and datam <1> was the RMS value. From the plotted data for Gain and Gain2, the 3-dB bandwidth was found to be th e same for each plot. The Quality Factor was also calculated. The necessary calculati ons and results are presented in Figure 3.7.

PAGE 39

261. High,Low 3dB frequency and Center Frequency f h 100.8103 f l 99.2103 f center 10010 3 2. Calculated 3dB Bandwidth B 3dB f h f l B 3dB 1.6103 3. Calculated Quality Factor Q f center B 3dB Q62.5 Figure 3.7: Analysis of the Quality Factor In the implementation, the Quality factor improved by a factor of two compared to the simulated value. Overall the sec ond order, narrow-band, bandpass Filter design met specifications and was satisfactory for the Arterial Perfusion Detection Method. 3.1.4 Constant Current Source An ideal Current Source is a device that maintains a constant current through a circuit, regardless of the load impedance. Its design can be achieved through the use of transistors. A transistor is a device that can amplify and produce an output signal with more power than its input signal. Circuit c onfigurations such as the current mirror and the push-pull configuration are used to impl ement a constant current source. For the Arterial Perfusion Detection, a transistor circuit similar to the push-pull configuration was designed. Instead of using the usual NPN/ PNP pair of transist ors, the reverse was utilized. One of the major benefits of th e chosen configuratio n was that its output impedance was much larger than most configur ations. This effect was achieved since the collector of a transistor exhibits a much larger impedance than the emitter. The larger impedance was necessary to prevent signal lo ading when the Arterial Perfusion device

PAGE 40

27 was connected to the skin, which can exhibit large impedances at certain frequencies. Figure 3.8 presents a schematic of the constant current source designe d for this research. 0 R11 {OHMS} G -VEE Rs 50 Q2 Q2N3904 0 Q1 Q2N3906 R4 10.48k C2 .01u C1 1u Vin R1 23k G V3 FREQ = 100k VAMPL = 5 VOFF = 0 R6 23k C4 1u 0 R2 23k R5 23k R7 10.48k VEE Figure 3.8: Schematic Diagram of the Constant Current Source 3.1.4.1 Circuit Analysis The design of the constant current source had to satis fy the conditions: 500uA quiescent current, 100 kHz Operating frequency, High Output Impedance, 12 Volts supply.

PAGE 41

28 The current source operating frequency and qui escent current were chosen to coincide with the electrical properties of human tissue. At high frequencies, the signal penetrates the deeper tissue layers under the stratum corn eum, which yields a deeper analysis of the body under study. The Bipolar Junction Transi stor was biased in acco rdance with the theory presented in reference [4]. PNP Transistor Analysis : 0 R11 {OHMS} 0 Q1 Q2N3906 R4 10.48k C1 1u Vin R1 23k 0 C4 1u R2 23k I2 INPN R8 VEE Figure 3.9: Top Half of the Constant Current Source The quiescent current was specified to be 500uA. Using the formula relating collector current to emitter curren t of a transistor the emitter, IE, was found to be 502uA with a DC beta value of 198. The base voltage was chosen to be a fraction of the supply voltage, (6 Volts), which en abled the computation of RE, which yielded a value of 10.48 kOhm. The biasing design equations a nd results are presented in Figure 3.10.

PAGE 42

29R E 10.48k V BB I B R BB V BE I E R E solveR E 10488.944723618090452 6. R BB 1.15104 R BB R B2 R B1 R B2 R B1 5. R B1 23k V BB V EE R B2 R B2 R B1 solveR B1 23000 4. I B 2.525106 I B I E I c 3.I E 5.025104 I E I c 2.I c 500u 1. Calculation: Arbitrary valueR B2 23103 Base & Emiiter junction voltageV BE .7 Used 1/2 of power supply voltageV BB 6 Power SupplyV EE 12 Obtained form curve tracer 198 Chosen: Figure 3.10: Analysis of the Top Half of the Constant Current Source Since the two transistors were arranged in a manner complimentary to each other, the resistor values associated with the NPN transistor were the same as the ones calculated for the PNP transistor.

PAGE 43

30 NPN Transistor Analysis : -VEE Q2 Q2N3904 0 C2 .01u R6 23k 0 R5 23k R7 10.48k I2 INPN 0 R8 Figure 3.11: Bottom Half of the Current Source 3.1.4.2 PSPICE Software Simulation The constant current source was built in PSPICE and simulated. Two main objectives were considered using its model. The first was to analyze and compare the DC node voltages and the transistor currents, wh ich are shown in Appendix 3. The second was a check for linearity of the current sour ce when different loads were applied. The values for the DC node voltages and currents calculated and simulated are presented in Table 3. The simulation results were very similar to those calculated. Table 3.2: DC Characteris tics for the BJT Simulation DC Characteristics Calculated Simulated Collector Current 500uA 502.9uA Emitter Current 502.5uA 505.1uA Base Voltage 6V 6.025V

PAGE 44

31 Linearity of the current source: The circuit was simulated using a parametr ic sweep of the output load that was varied from 100 Ohm to 1 kOhm. The results are presented in the plots of Figures 3.12 and 3.13. Figure 3.12: Output Voltage with Fluctuating Load Resistor Figure 3.13: Output Current with Fluctuating Load Resistor Figure 3.12 demonstrates how the output volta ge varied as the load was varied. As the load changed the network adjusted its voltage to maintain a constant current. Figure 3.13 displays the current th rough the load as the load wa s varied. This plot also confirms that the current remained constant when the load was varied. The simulation

PAGE 45

32 demonstrated that the constant current sour ce was independent of the variation in the output load. 3.1.4.3 Implementation Results The constant current source was physically implemented and its linearity investigated in order to determine if the ci rcuit was adequate for the Arterial Detection Method. In order to verify its linearity, the test procedure presented in Appendix 4 was developed. The data obtained is presented in Figure 3.14. 1 10 41.5 10 42 10 42.5 10 43 10 43.5 10 44 10 44.5 10 45 10 40 0.1 0.2 0.3 0.4 0.5 CurrentVolatge0.492 9.902103 v0 v1 v2 v3 v4 v5 v6 4.968104 10105 I R1 I R2 I R3 I R4 I R5 I R6 I R7 Figure 3.14: Measured Data from the Multi-Meter The plots presented in Figure 3.13 demonstr ate that the voltage across the varying resistor was proportional to the current thr ough the varying resistor The design of the constant current source was successfully impl emented and exhibited the linearity required for its use in the Arterial Perfusion Detection Method. 3.1.5 Emitter Follower The emitter follower was designed to act as a buffer between the differential amplifier and the two inner electrodes, which were placed on the skin. Its characteristics allows for the connection of a source having high resistance, Rs, to a load with relatively

PAGE 46

33 low resistance, [4]. Another useful aspect of the emitter follower is that its gain is unity. Furthermore, the input coupli ng capacitor with the bias resi stor at the base of the transistor forms a high pass filter and, sin ce the output is taken across the emitter, no phase inversion in signal is generated. Figure 3.15 presents the schematic for the emitter follower designed during this research. C2 1u G C1 1u Q1 Q2N3904 V2 FREQ = 100k VAMPL = 1 VOFF = 0 V G G R1 10k G R2 20k R3 250 Rskin 500 VCC R4 680 Figure 3.15: Emitter Follower 3.1.5.1 Circuit Analysis The emitter follower was designed to comply with the specifications: High Input Impedance, Input RC configuration yi eld >10 Hz 3-dB point, 12 Volts Power Supply, VE = 5 Volts. The high input impedance was desired to prevent any signal loading when the emitter follower was connected to the subject’ s skin. Additionally, the cut-off frequency was chosen to be greater than 10 Hz in orde r to remove the large DC offset due to the

PAGE 47

34 equipotential at the electrode-skin contact. The calculation of the input and output impedances as well as the bias resistor values for the emitter follower is presented in Appendix 5. 3.1.5.2 PSPICE Software Simulation The Emitter Follower was built and simulate d with PSPICE. The objective of the simulation was to compare the calculated node voltages and currents to the simulation values. In addition, the frequency response of the design was examined to determine whether the specified 3-dB point was achie ved. The emitter follower circuit schematic depicting node voltages and branch currents is presented in Figure 3.16. DC Node Voltages and Transistor Currents: C1 1u G 0V 6.983V 0V Q1 Q2N3904 152.5uA 24.76mA -24.91mA R1 10k 501.7uA G 6.227V R2 20k 349.2uA 12.00V 0V G V R3 250 24.91mA R4 680 0A G V2 FREQ = 100k VAMPL = 1 VOFF = 0 0A Rskin 500 VCC C2 1u Figure 3.16: DC Node Voltages and Currents Table 4 presents the simulated and calculat ed values for the design. These values show that the transistor’s simulated values cl osely agreed with the transistor’s calculated values. Table 3.3 presents the operating volta ges and currents for th e transistor in the emitter follower circuit.

PAGE 48

35 Table 3.3: Transistor Oper ating Currents and Voltages Calculated Simulated VB 8V 6.983 V VE 7.3V 6.227V IB 154uA 152.5uA IE 25.09mA 24.91mA IC 24.94mA 24.76mA The frequency response of the emitter follower was investigated in order to demonstrate compliance with the desired spec ification. Figure 3.17 presents a response curve with a 3-dB point at 238 Hz for a highpass filter. Frequency 1.0mHz 10mHz 100mHz 1.0Hz 10Hz 100Hz 1.0KHz 10KHz 100KHz 1.0MHz V(R4:2) 0V 0.5V 1.0V (238.755,704.763m) Figure 3.17: Frequency Response of the Emitter Follower The design of the emitter follower exhibited a voltage gain of approximately unity, a low output resistance a nd a high input resistance. Th ese results were well suited for its application as a buffer in th e Arterial Perfusion Detection Method.

PAGE 49

36 3.1.6 BJT Differential Amplifier and Buffer One of the processing tools of the Arte rial Perfusion Dete ction Method is the differential amplifier. The differential amp lifier is commonly used for the amplification of the difference voltage between two input si gnals such as the voltage signals retrieved by the two inner electrodes of the four electrode system. The differential amplifier designed supplied a very good common mode re jection, (CMRR), of 87dB and provided a gain of 38V/V. The output for the differen tial amplifier can be taken differentially or single-ended. The Arterial Perfusion circuit required a si ngle-ended output, which was subsequently buffered. Figure 3.18 presents th e schematic for the differential amplifier and buffer designed during this research. G R31 10k Q5 Q2N3904 Q4 Q2N3904 R39 30k C2 1u Q6 Q2N3904 V++ C27 .1u C23 .1u R29 1k C1 1u R42 680 C26 1u V-R41 250 R3 680 R40 60k C24 .1u R4 680 C28 1u G G V-G G G G V++ G G G G G V33 FREQ = 100k VAMPL = .5 VOFF = 0 G 0 G V2 12 C25 .1u V++ Q3 Q2N3904 R45 20k G R30 1.25k R32 7.5k V3 FREQ = 100k VAMPL = .5 VOFF = 0 R28 1k V1 12 Figure 3.18: Schematic Diagram of the BJT Differential Amplifier and Buffer

PAGE 50

37 3.1.6.1 Circuit Analysis The differential amplifier was designed to meet the specifications: High CMRR to remove unma ted common mode voltages, Max 2.5mA operating quiescent current on Q1 and Q2, 12 Volts Power Supply. The analysis of this design was performed us ing the equations in, [4]. The specified emitter current could not exceed 2.5mA. Therefor e, the current source that was used to bias the differential amplifier was also used to bias the collector current of the emitter follower. The single-ended output was capacitive ly coupled with resistor to ground. The coupling circuit formed a highpass filter, whic h produced a large DC offset at the output node. The overall single ended output gain of the design without the buffer yielded the results presented in Figure 3.19. From PSPICE SImulation:I c 2.38103 g m .0893 R c 1k A cm 1.820710 3 Gain and CMRR calculation: 1. Gain 1 2 g m R c Gain44.65 2. CMRR20log Gain A cm CMRR87.792 dB Figure 3.19: Gain and CMRR Analysis of the Differential Amplifier

PAGE 51

38 3.1.6.2 PSPICE Software Simulation The Differential Amplifier was built and simulated in PSPICE. The purpose of the simulation was to verify the operation of the circuit and to perform a frequency analysis in order to determine its overall gain at the singleended output. Figure 3.19 presents a schematic of the differential amp lifier that depicts the DC current values obtained for the various branches of the circui t. The collector current of the differential amplifier was slightly less than the specified value. The current source connected to the emitters of the differential amplifier shows that proper biasing was provided since the collector currents were approximately the total emitter currents of the current source. R3 680 14.54uA R28 1k 2.380mA C2 1u G V-R31 10k 673.2uA G C24 .1u R32 7.5k 702.4uA G Q5 Q2N3904 29.12uA 4.792mA -4.821mA G G V++ V3 FREQ = 100k VAMPL = 1 VOFF = 0 0A Q3 Q2N3904 14.54uA 2.380mA -2.394mA R45 20k 0A G Q4 Q2N3904 14.54uA 2.380mA -2.394mA C25 .1u R44 500k 3.477uA R30 1.25k 4.821mA R4 680 14.54uA C1 1u R29 1k 2.380mA C23 .1u G G G V4 FREQ = 100k VAMPL = .990 VOFF = 0 0A Figure 3.20: DC Currents of the Simulated Differential Amplifier

PAGE 52

39 In order to verify the frequency response of the differential am plifier one of the inputs was set to ground and a frequency res ponse simulation performed. The results of the frequency response simulation are presen ted in Figure 3.21. The maximum gain of the differential amplifier circuit was 38.079V/V with the 3db point at 35V/V. Frequency 1.0mHz 10mHz 100mHz 1.0Hz 10Hz 100Hz 1.0KHz 10KHz 100KHz 1.0MHz V(R45:2) 0V 10V 20V 30V 40V (100.000K,38.079) Figure 3.21: Gain Response of the Differential Amplifier Another factor requiring consid eration was that the electr odes, applied to the skin, could have a separation of no more than 2 inches. The electrode separation could not yield a differential voltage of more than 50mV. This lim it was important in order to insure that the differential amplifier would not saturate. The desi gn of the Differential Amplifier was successfully performed and implemented in the Arterial Perfusion Detection Method. 3.1.7 Second Order Phase Shifter with Vo ltage Divider and Op-Amp Buffer The requirements that the signal must be in phase with the re ference signal, be less than 800mV and be buffered before be ing introduced to the AD630 for synchronous

PAGE 53

40 detection were additional criter ia of the Arterial Perfusion Detection Method that had to be met. To accomplish these tasks an allp ass phase shifter employed as a Phase Lag circuit, voltage divi der and buffer were implemented. An all-pass phase shifter is a circuit that will pass all signal frequencies without attenuation but with varying phase shifts, [6]. This allpass filter, voltage di vider and circuit is presented in Figure 3.22. U11 AD817/AD 3 2 7 4 6 + -V+ V-OUT V U9 AD817/AD 3 2 7 4 6 + -V+ V-OUT R50 1.5k G G R51 5.11k U10 AD817/AD 3 2 7 4 6 + -V+ V-OUT Rvariable2 {OHMS} V-R49 1.5k G G V++ V++ R46 1.5k C33 .1u G C34 .1u G R47 1.5k C36 .1u G G C31 .1u C35 .1u C29 .001u V-C30 .1u V-R52 1k R53 302 C32 .001u R48 1.5k R54 302 G V3 FREQ = 100k VAMPL = 5 VOFF = 0 V++ 0 Figure 3.22: Schematic Diagram of the Second Order All-Pass Filter, Voltage Divider and Buffer The voltage divider circuit that provided a fraction of the input voltage was simply based on the combination of two series resistors. The buffer, which was an opamp version of the emitter follower, was used to provide a high input impedance and low output impedance with unity gain. 3.1.7.1 Circuit Analysis The design of the second order allpass phase circuit had to be able to provide a phase lag that was adjustable from 0o to 90o. Using the methodology in, [6], the values for the necessary components were calculate d. Though the method in, [6], states that the

PAGE 54

41 allpass circuit be designed by first determining resistors R1 and R2, the latter was performed by choosing the capacitor value fi rst and working backwards. For a phase shift of 90o, the calculation procedure is illustrated in Figure 3.23. Given: C.001u f100k R 1 3k 1. C 1 2 f R 3 solveR 3 1591.5494309189533577 R 3 1.591 k 2. R 3 R 1 R 2 R 1 R 2 solveR 2 3387.5088715400993612 R 2 3.387 k Figure 3.23: Calculation of the Com ponent Values for a Phase Shifter The design required a second order phase shifter. Simply cascading two phase shifters produced the required second order allpass phase shifter. The voltage divider design equation is given by equation (5). Voltage Divider Equation V out R 2 R 1 R 2 V in (5) Equation (5) yielded the required step down voltage. The Op-Amp Buffer was obtained by using the Analog Devices component IC AD817. Further information about this integrated circuit can be f ound in the literature, [14]. 3.1.7.2 PSPICE Software Simulation The overall circuit was built and simu lated in PSPICE. The purpose of the simulation was to verify that a 90o phase shift by the first phase shifter was obtained when R3 of the design was set to its calculated va lue. Since the all-pass filter was second order, a total of 180o should be obtained at the output of the second phase shifter.

PAGE 55

42 Figure 3.24: Second Order Allpass Phase Shifter Simulation Results Figure 3.24 shows that the first phase sh ifter provided a phase difference of 86.04o and the second phase shifter provided an additional phase difference of 86.76o. The total phase shift for th e allpass circuit was 172.8o. The simulation demonstrated that the second order allpass phase shifter satisfied the design cr iteria. The design equations and the results of their us e are presented in Figure 3.25. Given: Ref15.015u PS 1 17.405u PS 2 19.815u Period of Reference waveform Period20.01u10.01u Phase Difference: Reference and first Phase Shifter PS 1 Ref Period 360 86.04 Phase Difference: Reference and second Phase Shifter PS 2 Ref Period 360 172.8 Figure 3.25: Verification of the Phase-Shifter Design The phase shifter has a second-or der transfer function given by: Hs ()k s2 o Q o2 s2 o Q o2 where o 1 R 1 R 2 C 1 C 2 Q R 1 R 2 C 1 C 2 R 1 C 1 R 2 C 2 (6)

PAGE 56

43 Application of the transfer function to the source code in Appendix 1provided the results presented in Figure 3.26. The top graph pr esents the pole/zero plot. It is observed that the poles and zeroes disp layed mirror-image symmetry with respect to the jw axis, which is required of a well de signed allpass phase shifter. Figure 3.26: Allpass Phase Shifter Pole/Zero Plots The theoretical and simulated circui t was properly designed and met the specifications for its use in the Ar terial Perfusion Detection Method. 3.1.8 AD630 Integrated Circuit The AD630 IC was the most instrumental pa rt in the Arterial Perfusion Detection Method. Its purpose was to detect the amplit ude modulation imposed on of the retrieved differential voltage signals from the inne r-electrodes by synchronous detection. The Analog Devices AD630 IC is a high precisi on balanced modulator/demodulator that

PAGE 57

44 incorporates signal processing applications such as balanced modulation, demodulation, synchronous detection, phase detection, quadrat ure detection, phase -sensitive detection, lock-in amplification and square wave multiplication. Figure 3.27: Functional Block Diagram of the AD630 3.1.8.1 Synchronous Detection Synchronous detection works well for frequencies up to several mega-Hertz. The synchronous detection or homodyne detec tion method dynamically combines range, accuracy and speed. This design implemented th e rectification by inversion of the output during alternate cycles, [5]. To obtain optim um performance, a minimal noise signal at the same frequency and phase as the carrier of referenced amplitude modulated signal being detected was required. In the Arterial Perfusion circuit, the s ynchronous detection circuit was utilized for the detection of the signal th at was amplitude modulated by th e arterial blood flow in the

PAGE 58

45 limbs. The resulting output of the chip wa s a signal correspondi ng to the area under the curve under study. 3.1.8.2 PSPICE Software Simulation The simulation test was performed to ch aracterize the AD630 that was used for synchronous detection in the Ar terial Perfusion Hardware. In order to characterize the AD630, the design of a balanced modulator shown in the part’s data sh eet was built, [18]. The operational schematic for the AD630 is presented in Figure 3.28. Figure 3.28: AD630 as a Gain-of-One Balanced Modulator Pins 6, 5, 4 and 3 of the AD630 chip were not connected since their purpose of adjusting the DC offset was not the primary focus of the test. A low-pass filter was added at the output of the modulator to remove any higher frequency components. Figure 3.29 presents a schematic of the simula tion circuit that util ized the AD630 as a component. Initially, two inputs to the AD 630 were generated. One input was a 100 kHz square wave and the other was a 10 kHz sine wa ve. In order to replicate the data found in the AD630 data sheet the amplitudes of the square wave and sine wave were set at 1Volt.

PAGE 59

46 Vee V3 TD = 0 TF = .01u PW = 5u PER = 10u V1 = -5 TR = .01u V2 = 5 Vee VCC G V C1 .01u VCC 0 U2 AD630J/AD 2 7 8 9 11 14 15 16 17 18 19 20 13 10 12 1 3 4 5 6 CHA+ STAT -VS SELB +VS RB RF RA RINB CHB+ CHBCHAVOUT SELA COMP RINA DOA1 DOA2 COA1 COA2 Vsin C2 .01u V4 FREQ = 10k VAMPL = 5 VOFF = 0 G Vsin R7 200 Vpulse R2 1k G R5 1MEG Vee U3 AD817/AD 3 2 7 4 6 + -V+ V-OUT C3 .001u R1 50 G Vpulse R9 500 V2 12 G R8 200 0 VCC V1 12 0 C4 .001u 0 G R3 1MEG R4 1MEG G G Figure 3.29: PSPICE Schematic of the Simulation Figures 3.30, 3.31 and 3.32 present the simulate d waves and the output waveform of the AD630 Balanced Modulator. Time 0s 20us 40us 60us 80us 100us 120us 140us 160us 180us 200us V(Vpulse) -1.0V 0V 1.0V Figure 3.30: PSPICE Generate d 100 kHz Square Wave

PAGE 60

47 Time 0s 20us 40us 60us 80us 100us 120us 140us 160us 180us 200us V(Vsin) -1.0V 0V 1.0V Figure 3.31: PSPICE Generated 10 kHz Sine Wave Time 0s 20us 40us 60us 80us 100us 120us 140us 160us 180us 200us V(U3:OUT) -1.0V 0V 1.0V Figure 3.32: AD630 Output Simulation 3.1.8.3 Implementation The implementation of the AD630 Bala nced Modulator followed the same procedure that was performed in the PSPICE simulation except for the inclusion of a lowpass filter. A square wave with a fr equency of 10 kHz and amplitude of 1V was generated with a function generator. The 100 kHz square wave and the 10 kHz sine wave are depicted graphically in Fi gures 3.33 and 3.34 respectively.

PAGE 61

48 Figure 3.33: 100 KHz Square Wave Figure 3.34: Ten Kilohertz, (10 KHz), Sine Wave Figure 3.35: Implementation Output from the AD630 Analyses of the simulation data, implemen tation test data and the data from the specification sheet, yields the conclusion that the chip behaved as required when it was connected in the circuit. The only observable difference was the amplitude of the generated waveform. The specification asked for a 5V amplitude while the experiment was performed with a 1V amplitude. The data sheet results, AD630 output simulation

PAGE 62

49 results and the implementation output from the AD630 are presented in Figures 3.36, 3.37 and 3.38 respectively. Figure 3.36: Data Sheet Results Time 0s 20us 40us 60us 80us 100us 120us 140us 160us 180us 200us V(U3:OUT) -1.0V 0V 1.0V Figure 3.37: AD630 Simulation Output Figure 3.38: AD630 Implementation Output 3.1.9 Sallen & Key Second Order Active Highpass Butterworth Filter The output of the AD630 synchronous detect or generated a large DC offset as well as frequencies twice the cen ter frequency of its input. In order to remove the large

PAGE 63

50 DC offset a highpass filter was employed. A nother reason to filter the AD630 DC offset was that the frequency of blood flow is in th e range from 500mHz to 3Hz. Therefore, the choice a 3-dB cut off point for a highpass filter of less than 500mHz was required. The High-Pass filter incorporated in the Arterial Perfusion Detection Method was a Sallen & Key second order activ e highpass Butterworth filter. This filter is a two pole filter that exhibits a maximally flat response. The schematic for the filter designed during this research is presented in Figure 3.39. G R19 100 V++ R18 332k G R17 332k G V3 FREQ = 100k VAMPL = 1 VOFF = 0 C4 .1u G C13 .1u U4 AD817/AD 3 2 7 4 6 + -V+ V-OUT C12 6.8u V-C11 6.8u Figure 3.39: Sallen & Key Highpass Filter This circuit belongs to the family of Sallen & Key filters, which are named after their original designers. The design equations of this filter are very straightforward and easy to use for fast implementation.

PAGE 64

51 3.1.9.1 Circuit Analysis The High-Pass filter was designe d to meet the specifications: Fcutoff = < 1 Hz, Non-inverting input, 12 Volts Power Supply. The design equations listed in, [4], were follo wed to achieve the above criteria and the necessary resistor values were obtained afte r first choosing the value for the capacitor. Fig 3.40 illustrates the an alysis of the filter. Figure 3.40: Analysis of the High-Pass Filter The filter arrangement required the feedback loop to be attached to the non-inverting terminal of the op-amp. Doing so produced the high input impedance requirement cited by the specification. 3.1.9.2 PSPICE Software Simulation The filter was designed and simulated us ing PSPICE software. The objective of this simulation was to verify whether the ca lculated 3-dB cut-off point was obtained and to evaluate the position of the poles and zeroes of the filter to ensure stability. The first

PAGE 65

52 simulation performed was the frequency respons e of the filter. Figure 3.41 illustrates the response with a 3-dB cut-off of 105.5 mHz. This cut-off point satisfied the design criteria. Figure 3.41: Simulated Frequency Response of the Highpass Filter After the frequency response was acquire d the pole and zero placements were evaluated. To evaluate the poles and zeroes of the filter the tran sfer function, equation (7), was programmed into Matlab and analyzed. Hs () KS2R 1 R 2 C 1 C 2 s2R 1 R 2 C 1 C 2 sR 2 C 2 R 2 C 1 R 1 C 2 1K () 1 (7) Figure 3.42 shows that the locations of the poles of the transfer function are distant from the jw axis, which signifies that the filter was stable. The 3-dB gain of the filter showed a frequency response similar to the one obt ained during simulation. This outcome demonstrated that the High-Pass Filter wa s properly designed and met specifications.

PAGE 66

53 Figure 3.42: Pole/Zero and Gain Response of the Highpass Filter 3.1.10 AD620 Integrated Circuit The AD620 is a monolithic instrumentation amplifier based on a modification of the classic three op-amp approach. Absolute va lue trimming allows the user to accurately program the gain, with only one resistor, to 0.15% at a gain equal to 100, [7]. Figure 3.43 presents the AD620’a func tional block diagram. The stage following the highpass filter was the AD620, which was utilized for amplification of the signal. Normally signals are filtered first and then amplified. However, for this design the optimum pe rformance was obtained by highpass filtering, amplifying and then lowpass filtering the waveform.

PAGE 67

54 Figure 3.43: Functional Bloc k Diagram of the AD620, [7] The AD620 is meant to be used as a differe ntial amplifier. However, the Arterial Perfusion Detection circuit used this amplif ier for its gain characteristics, which were capable of providing a gain of 1000 by setting the inverting input to ground. With the amplitude of the desired signal unknown and exp ected to be less than a millivolt, setting the gain of the amplifier to 1000 di d not cause saturation at the output. 3.1.10.1 Circuit Analysis The equation for setting the gain of th e amplifier is published in, [7], and presented here as equations (8): Gain 49.4k R G 1 R G 49.4k Gain1 (8) To obtain the necessary resistor value for a gain of a thousand, the second formula presented in equation (8) was used. Figure 3.44 presents the value for the gain setting resistor.

PAGE 68

55 Figure 3.44: Analysis of the Gain Resistor 3.1.10.2 PSPICE Software Simulation The amplifier was constructed and simula ted in PSPICE. Figure 3.45 presents the schematic that was constructed in PSPICE. R3 100 R4 500 V-G R2 4.99k R1 51 G U1 AD620/AD 6 3 2 5 7 4 1 8 OUT + -REFV+ V-RG1 RG2 V++ V3 FREQ = 1k VAMPL = .001 VOFF = 0 G G Figure 3.45: PSPICE Schematic of the AD620 A standard value of 51 ohms was implemente d for the gain resistor. A sine wave with amplitude of 0.001mV was used for the si mulation input. The output of the circuit yielded a waveform with peak amplitude of 1 Volt. This outcome demonstrated that a gain of 1000 was effectively obtained by th e configuration of the AD620 integrated circuit. The output of the AD620 duri ng simulation is presented in Figure 3.46.

PAGE 69

56 Figure 3.46: AD620 IC Simulation Output 3.1.11 Third Order Passive RC Filter The last stage of the Arte rial Perfusion circuit wa s a second order passive RC filter. Several different filters, includ ing both active and passive designs were implemented and tested during this research. However, the best results were obtained by a second order passive filter. The schematic for the best filter designed during this research, which was passive, is presented in Figure 3.47. R5 3.5k R7 3.5k 0 C4 3.3u C5 3.3u C6 3.3u V4 FREQ = .8 VAMPL = 1 VOFF = 0 0 0 R6 3.5k 0 Figure 3.47: Third Order Passive RC Filter The literature review indicated that the frequency at which blood flows is approximately one hertz. Therefore, the f ilter’s 3-dB cut-off frequency was designed to be no more than five hertz so that al l cardiac and higher order frequencies were attenuated.

PAGE 70

57 3.1.11.1 Circuit Analysis The evaluation of the filter components for the lowpass filter used the basic design equation in, [4]. Figure 3.48 presen ts the overall analysis of the design. Given: C3.3106 f c 30 Calculation of R: f c 1 2 R C solveR R1.6103 Figure 3.48: Analysis of RC Filter Components 3.1.11.2 PSPICE Software Simulation The design was constructed in PSPICE a nd the simulated frequency response of the filter was observed. At first, the calculated design values were used for the simulation. However, the required response wa s not achieved. Performance of a Sweep Simulation of the resistor value allowed for the selection of a frequency response that corresponded to the overall specification. Fi gure 3.49 presents the response chosen for the design of the Third Order Passive Lowpass RC Filter. Successful testing and implementation of the lowpass filter completed the design for the Arterial Perfusion Detector.

PAGE 71

58 Figure 3.49: Passive Lowpa ss Filter Frequency Response

PAGE 72

59 CHAPTER 4 SOFTWARE DESCRIPTION 4.1 Software Description The software design used to process the data obtained by the Arterial Perfusion circuit will be discussed in this chapter. Th e use of a data acquisition card and a personal computer interface will be presented in deta il. The personal interface that was designed allows the user to view the acquired data in real time as well as perform mathematical operations such as performing the derivative on the retrieved data. 4.1.1 National Instruments LabView LabView is a graphical programming langua ge produced by National Instruments. LabView is used primarily for signal processi ng of signals ranging from Instrumentation Control to Bio-Medical applic ations. The LabView graphi cal dataflow language and block diagram approach to problem solving naturally represent the flow of data. LabView provides the ability to intuitively map user interface controls to allow convenient viewing and modifica tion of the data. In additio n LabView provides the same capability for control inputs, [10]. Figure 4.1 illustrates the programming style associated with the LabView graphical programming language.

PAGE 73

60 Figure 4.1: Illustration of the LabView Programming Style 4.1.2 Data Acquisition Data acquisition is the activity associated with obtaining and st oring data in real time from analog and digital based devices such as circuits and instrument used during and/or under experimentation. Its functionality generally in cludes a combination of PCbased measurement hardware and software to provide a flexible, user-defined, measurement system, [10]. Figure 4.2 presents a data acquisition system diagrammatically. Figure 4.2: Data Acquisition System, [10] The data Acquisition hardware used duri ng this research was the PCI-MIO-16E. Part of the E Series family of devices, the PC I data acquisition card is one of the fastest and the most precise multiplexed data acquisition devices produced by National Instruments. The PCI-MIO-16E card is ideal for applicatio ns such as continuous high-

PAGE 74

61 speed data logging, control applications and hi gh voltage signal or sensor measurements when used with NI signal conditioning, [ 10]. The full data specification for the PCIMIO-16E data acquisition card is presented in, [10]. 4.1.3 LabView Interface The software interface will illustrate the sampled output signal of the Arterial Perfusion Detection Circuit. The LabView interface developed for this research was comprised of four graphs. The first graph pr esents the sampled electro-cardiogram signal from the first analog input channel of th e data acquisition card. The second graph presents the sampled photo-plethysmography signal obtained from the second analog input channel of the data acquisition card. Th e third graph presents the sampled Arterial Perfusion Detection circuit out put signal and the fourth grap h presents the derivative of the output signal. Figure 4.3 presents the Arteri al Perfusion Detection software interface. Figure 4.3: Software Interface for the Arterial Perfusion Detection Circuit

PAGE 75

62 The derivative of the filtered signal was monitored so th at the rate of change of the signal could be examined when the ar m cuff was utilized during experimentation. A filter selection capability was also inco rporated in the interface in order to remove any unwanted signals or annoying noise A total of three digital filters were designed. The digital filter set was comprised of a 60Hz notch filter, a 60Hz Lowpass filter and a 30Hz Lowpass filter. The algorithm generated for the labView interface is presented in Figure 4.4. Figure 4.4: LabView Graphical Codes

PAGE 76

63 CHAPTER 5 TESTING AND RESULTS 5.1 Forearm Limb Experimentation The first site chosen for the Arterial Perfusion Detection Method experiment was close to the wrist where the radial artery re sides. This location is commonly used by medical personnel for the detection of the radial pulse. The pulse is the outcome of pressure wa ves moving through the arterial vessels. When the pulse is taken from the wrist, the index and middle fingers are placed on the radial artery, which is located on the wrist prox imal to the hand. Therefore, the same site was used for the placement of the electrode s during this research. The spacing of the electrodes was kept uniform and at a 5mm di stance from each other in order to have a greater degree of detection of the pulse. Fi gure 5.1 pictures the position of the test electrodes.

PAGE 77

64 Figure 5.1: Test Electrode Placements 5.1.1 Arm-Cuff: Deflated The first experimentation performed was an attempt to simply detect the arterial pulse with the arterial perfusion design a nd the photo-plethysmograph circuit. Figure 5.2 presents the measurement setup for the forear m limb arterial pulse detection experiment. Figure 5.2: Setup for the Forearm Limb Measurement

PAGE 78

65 In the picture presented in Figure 5.2, the electrodes are connected to the arterial perfusion detection circuitr y, on the red breadboard, via in sulated wires. The photoplethysmograph, built on the black breadboa rd, was connected to the middle finger. The photo-plethysmograph was also impl emented in the test. Capturing waveforms using both the arterial perf usion detection method and the photoplethysmograph facilitated the determination of whether the waveform obtained from the arterial perfusion detection circuit was synchronous with the photo-plethysmograph waveform. The waveform obtained by the ar terial perfusion detection method and the photo-plethysmograph are presented in Figure 5.3. Figure 5.3: Arterial Pulse from Arteri al Perfusion and Photo-Plethysmograph Study of the graphs presented in Figure 5.3 reveals that, in the ten second time frame of the test, both methods detected the same number of peaks, which was thirteen, (13), in numbers. This result demonstrates that the arterial perfusion circuit was functioning properly and was synchronous with the photo-plethysmograph signal.

PAGE 79

66 5.1.2 Arm-Cuff: Inflated and Released The next experimentation util ized an arm-cuff, which is a devise used to measure blood pressure. The arm-cuff utilized during this research is pictured in Figure 4.4. Figure 5.4: An Arm-Cuff The purpose of the arm-cuff was to stop arteri al blood flow into the forearm limb area. Applying this constriction fac ilitated a determination of wh ether the arterial perfusion method was detecting the absence of a pulse in the forearm, which would be identified by a change in impedance of the skin. Figure 5.5 shows the data obtained when the arm-cuff was deflated at the beginning of the simula tion, inflated to 180 mmHg for approximately 15 seconds and then released. Figure 5.5: Simulation with an Arm-Cuff

PAGE 80

67 Figure 5.5 reveals several important factor s. During the initial period, when the arm-cuff was deflated, a pulse was detected in both th e arterial perfusion detection method and the photo-plethysmograph, (PPG), gr aph. As the arm cuff was inflated to 180 mmHg, the signal amplitude of the PPG slow ly decreased while the arterial perfusion design waveform showed some distortion, wh ich arose due to the sensitivity of the design. Any movement of the measuring arm will cause this effect. Therefore, it is best to be still when performing a measurement. Both graphs show no significant pulse once the arm-cuff is fully inflated, which is expe cted due to cessation of arterial blood flow. Once the arm cuff was released a spike in th e Arterial Perfusion De tection waveform was obtained. Physiologically, the spike is due to dilation of the arteries in response to local signals from the temporarily obstructed, isch emia, tissue for more oxygen. This outcome can also be evaluated by obser ving the derivative or rate of change of the waveform. Figure 5.6 presents the original filtered ar terial perfusion detection signal and its derivative.

PAGE 81

68 Figure 5.6: Original and Derivative Wavefo rms of the APD Design The derivative presented in Figure 5.6 signi fies the rate of change of the signal over time. It is apparent that the spike ge nerated in the top graph by the release of the arm-cuff creates a large change in the derivative signal. This result demonstrates that the signal retrieved by the arterial perfusion co rresponds to the impedance changes between the two voltage electrodes when the amount of blood at the chosen limb site fluctuates. Effectively, this result may not be obtained if a different site had been chosen. For example, performing an experiment with the ar terial perfusion detec tion circuit will yield no impedance signal when the electrodes are pla ced on the skin where the muscles, flexor carpi radialis and palmaris longus, are locate d. This outcome results from the lack of contrast between the skin and the radial ar tery, which lies beneath the brachioradialis. Figure 5.7 presents the anatomy of the forearm limb.

PAGE 82

69 Figure 5.7: Anatomy of the Forearm, [19] 5.2 Ankle Limb Experimentation The final site chosen for the Arterial Perfusion Detection Method experiment was to in the ankle area where the posterior tibial artery resides. This location is commonly used by medical personnel for the detection of the posterior tibial pulse in the lower extremities. The posterior tibial artery lies along the ca lf between the soleus and the deeper muscles of the lower leg limb. The posterior tib ial artery continues into the sole of the foot passing behind the medial ma lleolus. In the sole of the foot the artery divides into the medial and lateral plantar arteries, which anastomose, (join), with each other and with the dorsalis pedis artery to supply the anterior foot and toes, [20]. Figure 5.8 pictures the experimental setup for th e Ankle Limb experiment.

PAGE 83

70 Figure 5.8: Ankle Limb Experimentation 5.2.1 Arm-Cuff: Deflated The first experiment performed was to simp ly detect the posterior tibial pulse with the arterial perfusion design and the photo-pl ethysmograph circuit. Figure 5.8 presents the measurement setup for the ankle limb arteri al pulse detection. The connection of the electrodes to the circui ts was essentially sim ilar to that mentioned in section 5.1.1 as well as the use of both circuits for capturing the waveforms. Figure 5.9 presents the signals obtained from both the arterial perfusion circuit and the photo-plethysmograph circuit.

PAGE 84

71 Figure 5.9: APD and Phot o-Plethysmograph Signals Study of the graphs presented in Figure 5.9 reveals that dur ing the four second time frame of the test, both methods detected the same number of base peaks, which was four, (4), in numbers. The decision to count the base peaks instead of the top peaks is due to the poor signal integr ity of the photo-plethysmograph. Several attempts were made to obtain a better result but the photoplethysmograph device was initially designed specifically for use with the index finger. Ne vertheless, this result demonstrates that the arterial perfusion circuit was functioning properly and was synchronous with the photoplethysmograph signal.

PAGE 85

72 5.2.2 Arm-Cuff: Inflated and Released The next experiment utilized the same ar m-cuff, which is pictured in Figure 5.4. The purpose of the arm-cuff was to stop arteri al blood flow into the ankle limb area. Applying this constriction facili tated the determination of wh ether the arterial perfusion method was detecting the absence of the posteri or tibial pulse in the ankle. Figure 5.10 presents the data obtained when the arm-cu ff was deflated at the beginning of the simulation, inflated to 180 mmHg and finally released. Figure 5.10: Simulation with the Arm Cuff Study of Figure 5.10 reveals several importa nt factors. During the initial 9 seconds that the arm cuff was deflated, a pul se was detected by the arterial perfusion detection method. The detected si gnal is presented in Figure 5.11.

PAGE 86

73 Figure 5.11: Initial Pulse of the Ankle Experiment As the arm-cuff was inflated to 180 mmHg the arterial perfusion design waveform showed some distortion. As previously me ntioned in section 5.2.1, any movement of the measuring ankle limb will cause this effect Once the arm-cuff was fully inflated the graph shows reduced amplitude for pulsatile impedance changes. The small pulsatile impedance that is still present indicates th at the arm cuff did not totally occlude the arterial perfusion in the calf. These small ch anges in the pulsatile impedance of the calf are presented in Figure 5.12. These changes oc curred due to the inability of the arm-cuff to occlude the lower limb vessels. A cuff wider than the diameter of the limb at the site of application must be used in order to occlude arterial perfusion to the site.

PAGE 87

74 Figure 5.12: Pulsatile Impedance w ith the Arm Cuff Fully Inflated When the arm-cuff was released a spike a nd the return of pulsatile impedance in the Arterial Perfusion Detection waveform we re obtained. Physiologically, the spike is due to dilation of the arteries in response to local signals from the temporarily obstructed, ischemia, tissue for more oxygen. This outco me can also be evaluated by observing the rate of change or derivative, of the wavefo rm. Figure 5.13 presents the original filtered arterial perfusion detecti on signal and its derivative. Figure 5.13: Original Filtered APD Signal and the Signals’Derivative

PAGE 88

75 The derivative waveform presented in Figure 5.13 signifies the rate of change of the signal over time. It is appare nt that the spike generated in the top graph by the release of the arm cuff creates a large change in th e derivative signal. Overall, this result demonstrates that the signal retrieved by th e arterial perfusion method corresponds to the impedance changes between the two voltages electrodes when the amount of blood at the chosen limb site fluctuates. Figure 5.14: Initial Pulsatile Changes and Their Derivative Figure 5.15: Thorax Impedance Curve, [9]

PAGE 89

76 A comparison of the shape of the deri vative signal obtained from the APD method and the shape of the derivative of a thorax impedance curve shows that they are very similar. The derivative curves are pr esented in Figures 5.13 and 5.15. This factor confirms that the derivative waveform obtained by the APD method was accurate. 5.2.3 Motion Artifacts This section examines the effects of limb movements on the arterial perfusion circuit output signal. As mentioned in th e previous sections a ny motion by the patient will degrade the integrity of the captured signa l. For this experiment the subject was asked to slightly move his big toe, less than 3 degrees in the upward direction, in order to simulate a muscle spasm. Figure 5.16 demons trates the resulting signal when the subject moved the big toe, slightly, at two different instances. Figure 5.16: Output Signal of the Arte rial Perfusion Circ uit with Movement

PAGE 90

77 Study of figure 5.16 reveals that duri ng the 6 second time frame the pulsatile impedance changes in the limb were initially present. When the subject moved the big toe slightly, the impedance si gnal underwent a large altera tion. This change can be observed during the 5 second to 10 second time frame in Figure 5.16. Figure 5.17: Effects of the Ini tial Movement of the Big Toe Study of Figure 5.17 reveals that a larger pulsatile impedance change was obtained. This outcome can be deducted from the concept that when a muscle is stretch, not compressed, an increase in blood will be de livered to that area. After the subject relaxed the big toe, the impedance signal was ag ain prominent. The subject was asked to move his big toe again and a similar motion ar tifact was obtained compared to the initial one. This result demonstrates that the arteri al perfusion detection method design is very

PAGE 91

78 sensitive to motion and that any movement w ill result in the disintegration of the desired impedance signal.

PAGE 92

79 CHAPTER 6 CONCLUSION AND RECOMMENDATIONS 6.1 Conclusions The main objective of this research wa s to develop and implement an Arterial Perfusion Detection Method th at utilized the fundamenta l circuitry of Synchronous Detection. The device was required to work in conjunction with 4electrodes placed on the skin and detect pulsatile impedance changes of the skin. This research was to employ analog technology and software tools in orde r to realize the coupl ed advantages of adaptability, ease of reconstruc tion and the potential for future improvements. In order to carry out these objectives it was necessary to implement a design that employed sound circuit design tenants. Moreover, the cr eation of a program was required that could capture the outputs of the hardware in real time, provide a user interface and perform the derivative on the data. Implementation and testing of the device and software interface developed during this research clearly demonstrated that it is possible to design devices that will effectively detect pulsatile impedances and ultimately help the healthcare industry minimize the occurrence of pressure ulcers.

PAGE 93

80 6.2 Recommendations The implementation of the Arterial Pe rfusion Detection Method by Synchronous Detection revealed key issues that should be considered for future investigations. Specific recommendations for futu re investigations include: The hardware design develop during this research should be implemented on a Printed Circuit Board, (PCB). This act wi ll enable the mobility of the device, which will increase the number and variety of tests that could be performed. The software tools should be implem ented on a Personal Data Assistant, (PDA), which could provide increased awareness and control by giving a patient the capability of self monitoring. A wireless data transmission protocol s hould be included so that changes in the patients’ condition could be logged and monitored by a nurse.

PAGE 94

81 REFERENCES 1. A. Ivorra, “Design Considerations For Optimum Impedance Probes With Planar Electrodes For Bioimpedance Measurements”, [Online ], IEEE Xplore, Available at http://xplqa.ieee.org/Xplore/ guesthome.jsp, (accessed 15 July 2005) 2. Medical Engineer, “Measurement of Blood Flow”, [Online], Available at http://clinical.medicalengineer.co.uk /Measurement+of+Blood+Flow.php, ( access 25 July 2005) 3. Tektronix, “PS2521G Programmable Power Supply”, [Online Image], Available at http://www.tek.com/site/ps/0,,3M15100-INTRO_EN,00.html, (accessed 1 September 2005) 4. Kenneth C. Smith, “Microelectronic Ci rcuits”, Fourth Edition, Oxford, New York, 1998 5. Paul Horowitz and Winfield Hill, “A rt of Electronics”, Second Edition, Cambridge University Press, University of Cambridge 6. David A. Bell, “Operational Amplifiers ”, Prentice Hall, Englewood Cliffs, New Jersey 7. Analog Devices, “Low Cost Low Power Instrumentation Amplifier AD620”, [Online], Available at http://www.analog.com/UploadedFil es/Data_Sheets/897653854AD620_g.pdf, (accessed 1October 2005) 8. S. Grimned, “Bioimpedance & Bioelectri city”, Academic Press, New York 9. Jaakko Malmivuo & Robert Plonsey, “B ioelectromagnetism Principles and Applications of Bioelectric and Biomagne tic Fields”, Oxford Press, New York, 1995 10. National Instrument, “LabView 7.1” [On line], Available at http://www.ni.com 11. National Instrument, “LabView 7.1”, [On line], Available at http://www.ni.com

PAGE 95

82 12. D.C. Walker, B.H. Brown, D.R. Hose and R.H. Smallwood, “ Modelling the electrical impedivity of normal and premalignant cervical tissue”, [Online], Available at http://www.dcs.shef.ac.uk/ ~rod/Publications/Walker, (accessed 1October 2005) 13. Jang-Zern Tsai, “Dependence of Apprent Resistance of Four Electrode probes on insertion depth”, IEEE Magazine, 2001 14. Clinical Practice Guidelines, “Pressure Ulcer”, [Online], Available at www.pva.org 15. A.D.A.M. Medical Illustration Team, “Ar eas where Bedsores occur”, [Online Image], Available at http://www.nlm.nih.gov/medlineplus/ency/imagepages/19091.htm, (accessed 1October 2005) 16. Moor Instruments, “Laser Doppler T echniques”, [Online Image], Available at http://www.moor.co.uk/pdf/drt4.pdf (accessed 1 October 2005) 17. Prof. Dr. Charles McWilliams, “Multilayere d-skin”, [Online Image], Available at http://www.electrodermology.com/biophys prop.htm, (accessed 1 October 2005) 18. Analog Devices, “Balanced Modulat or/Demodulator AD630”, [Online], Available at http://www.analog.com/UploadedFil es/Data_Sheets/106127811AD630_e.pdf, (accessed 25 September 2005) 19. Vcudumuzu Tan yal m, “Forearm”, [Online image], Available at ww.metu.edu.tr/ ~guney/bb/anatomy .html, (accesed 31 November 2005) 20. Editorial Team, “Pulses of Lower Limb”, [On Line], Available at http://www.gla.ac.uk/ibls/fab /tutorial/generic/sapulse .html#posterior, (accessed 31 October 2005) 21. C.H Riedel, “Non-Contact Measuremen t Of The Electrical Impedance Of Biological Tissue”, [On Line], at IEEE Xplore, Available at http://xplqa.ieee.org/Xplore/guesthome.jsp, (accessed 15 July 2005) 22. JoAnn Maklebust, “Pressure Ulcers” Third Edition, Springhouse, Pennsylvenia 23. Gerald B. Zelenock “ Clinical Ischemic Syndromes”, Mosby Company, Missouri 24. John K-J. Li, “The Arterial Circ ulation”, Human Press, New Jersey

PAGE 96

83 APPENDICES

PAGE 97

84 Appendix A Matlab Source Code for the Transfer Function Narrow-Band Bandpass R1 = 1.5e3; R2 = 1.5e3; R3 = 1.5e3; R5 = 1.5e3; R6 = 332e3; C4 = 1000e-12; C6 = 1000e-12; %%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%% Second Order Highpass R1 = 332e3; R2 = 300e3; C1 = 6.8e-6; C2 = 6.8e-6; K = 1; %%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%% Second Order allpass R1 = 3e3; R2 = 3e3; C1 = .001e-6; C2 = .001e-6; wo = 1/(sqrt(R1*R2*C1*C2)); Q = (sqrt(R1*R2*C1*C2))/(R1*C1+R2*C2); k = 1; %%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%% Transfer Function NBP B = [0 K/(C6*R6) 0]; A = [1 1/(C6*R6) R2/(C4*C6*R1*R3*R5)]; V = tf(B,A); %%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%% Transfer Function HP B = [K*(R1*R2*C1*C2) 0 0]; A = [R1*R2*C1*C2 R2*C2+ R2*C1+(R1*C2*(1-K)) 1]; V = tf(B,A) %%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%% Transfer Function AP B = k*[1 -wo/Q wo^2];

PAGE 98

85 Appendix A: (Continued) A = [1 wo/Q wo^2]; V = tf(B,A); %%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%% Pole/Zero Positions and Gain Analysis figure(44) subplot(2,1,1) title('Pole Zero Plot') zplane(B,A) % Determination of Zeros disp('zeros are') z = roots(B) % Determination of Poles p = roots(A) figure(22) zplane(z,p) title('Pole Zero Plot') subplot(2,1,2) % Frequency Response of the Active Filter num = [-80 0]; den = [1 5 400]; w = logspace(-10,15); h = freqs(B, A, w); f = w/(2*pi); mag = 20*log10(abs(h)); semilogx(f, mag), title('Magnitude response') xlabel('frequency in Hz'), axis([0 500e10 -200 0]) ylabel('Gain in dB') figure(3) Bode(B,A,w)

PAGE 99

86 Appendix B Matlab Source Code for the Arterial Perfusion Detection Method close all clear all clc load data_24_10_2005.mat; a = TEK00007; y = a(:,2); y = (y./1); x1 = a(:,1); x1 = x1 x1(1); a1 = TEK00006; yy = a1(:,2); yy = (yy./1); xx = a(:,1); xx = xx xx(1); figure(5) subplot(2,1,1) plot(x1,y) ylabel('Amplitude') % title('PPG Pulse Reading') subplot(2,1,2) plot(xx,yy) ylabel('Amplitude') xlabel('Time') % title('Filtered PPG Pulse' % Applying Lowpass filter [B,A] = butter(3,.02) figure(11) Y = filter(B,A,y) subplot(2,1,1) plot(x1,Y) ylabel('Amplitude') % title('Arterial Perfus ion Detection Method') Y1 = filter(B,A,yy) subplot(2,1,2) plot(xx,Y1) ylabel('Amplitude') xlabel('Time') title('Photo-Plethysmograph')

PAGE 100

87 Appendix B: (Continued) % close all; % clear all; % clc %load data_24_10_2005.mat; % x1 = TEK00006(:,1); y1 = TEK00006(:,2); % x2 = TEK00007(:,1); y2 = TEK00007(:,2); % Derivative for TEK00006 dx1 = diff(x1); dy1 = diff(Y1); d1 = dy1./dx1; n = length(d1); d1 = [d1;d1(n)]; % the last two lines have to be added because % the diff function generates an array that is % shorter that x1 and/or y1 % Derivative for TEK00007 dx2 = diff(xx); dy2 = diff(Y); d2 = dy2./dx2; n = length(d2); d2 = [d2;d2(n)]; figure(111) subplot(2,1,1) plot(x1,Y1) title('Photo-Plethysmograph signal') subplot(2,1,2) plot(x1,d1) title('derivative') figure(5555) subplot(2,1,1) plot(xx,Y) title('Arterial Perfusion Design signal') subplot(2,1,2) plot(xx,d2) title('derivative') xlabel('Time')

PAGE 101

88 Appendix C Constant Current Source DC Currents and Node Voltages V3 FREQ = 100k VAMPL = 5 VOFF = 0 0A V++ C2 .01u G C1 1u R1 23k 259.8uA G G V-Q2 Q2N3904 3.844uA 502.9uA -506.7uA -6.044V R2 23k 262.0uA C4 1u Q1 Q2N3906 -2.190uA -502.9uA 505.1uA R11 100 0A 0 0V 12.00V -12.00V 6.707V I 12.00V 0V R4 10.48k 505.1uA G 0V Rs 50 0A 6.025V 842.4mV R5 23k 262.8uA R7 10.48k 506.7uA -6.690V V R6 23k 258.9uA Vin Figure C.1: BJT Node Currents and Voltages

PAGE 102

89 Appendix D Trouble Shooting the AC Voltage to Current Converter 0 G G R4 10.48k G R5 23k 0 G VEE R7 10.48k C1 1u Q2 Q2N3904 R2 23k Rs 50 0 C4 1u -VEE R1 23k Q1 Q2N3906 V2 -12 C.1u G C2 .01u R11 {OHMS} R3 1k V3 FREQ = 100k VAMPL = 5 VOFF = 0 -VEE Vin V1 12 VEE 0 C+ .1u R6 23k Figure D.1: Constant Current Source The test procedure requires: An Oscilloscope to measure the voltage across the 1kohm resistor, R3, A Digital Multi-Meter to measure the voltage across a second variable resistor, R11, which is in series with R1 at the circuit output,

PAGE 103

90 Appendix D: (Continued) A function generator capable of pr oviding a sine wave of variable amplitude. Vary the amplitude of the constant frequency input signal over the inclusive range from 1 volt to 5 volts. With the amplitude of the input signal established vary the value of R11 in fixed increments from 100ohms to 1kohms. Record the values of the voltages across R1 and R11.

PAGE 104

91 Appendix E Design of the Emitter Follower Schematic: C2 1u G C1 1u Q1 Q2N3904 V2 FREQ = 100k VAMPL = 1 VOFF = 0 V G G R1 10k G R2 20k R3 250 Rskin 500 VCC R4 680 Figure E.1: Emitter Follower Schematic Small signal analysis: r(pi) Rskin beta(i(b) R(E) R(L) R1 R2 Figure E.2: Small Signal Analysis Chosen: (Beta obtained from PSPICE) 162 R 1 10103 1 R 2 20103 R L 680 R E 25 0 R 1 10103 Essentially R1 and R2 are in parallel as well as R(E) and R(L): R l R L R E R L R E R B R 1 R 2 R 1 R 2 R B 6.667103 E B C

PAGE 105

92 Appendix E: (Continued) Calculation: V B 12 R 2 R 2 R 1 V B 8 I B V B .7 R B 1 R E I B 1.54104 I E 1 I B I E 0.02509 I c I E I c 0.02494 V CE 12I E R E V CE 5.726 r .026 I c r 168.881 Voltage Gain: A v 1 R l r 1 R l A v 0.994 Input Impedance: Output Impdeance: Z it r 1 R l R s 1 1 500 1 R 1 1 R 2 R s 465.116 Z i 1 1 R B 1 Z it Z i 5.453103 Z o 11 () R s r 1 R E Z o 3.89


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001912823
003 fts
005 20071012094221.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071012s2005 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001385
040
FHM
c FHM
035
(OCoLC)174113287
049
FHMM
090
TK145 (ONLINE)
1 100
Prevot, Yohan.
0 245
Arterial perfusion detection method by synchronous detection
h [electronic resource] /
by Yohan Prevot.
260
[Tampa, Fla] :
b University of South Florida,
2005.
3 520
ABSTRACT: The pressure ulcer is a well-known clinical problem that has plagued many patients in acute-care hospitals and chronic-care facilities. The pressure ulcer has the potential to diminish the quality of a patient's life by hindering the person's physical and emotional well-being. In addition, pressure ulcers are a high-cost problem. Past studies have shown that costs related to the treatment of pressure ulcers have reached 1.335 billion dollars a year in the United States alone. A pressure ulcer is defined as a lesion created by unrelieved pressure, which causes tissue ischemia and subsequently damages the underlying tissue. This sequence of events is mainly centered on ischemia. Ischemia is a condition created by an insufficient flow of blood to an organ or part of an organ such as the skin. The outcome of ischemia is cell death at the tissue level, which is commonly termed necrosis. In the past, researchers employed several different non-invasive techniques in order t o detect changes in the condition of human skin when blood flow was restricted. The two most commonly used practices were Laser Doppler Velocimetry and Continuous Wave Ultrasound. Laser Doppler Velocimeter is used to measure cutaneous blood flow in a study region. The moving red blood cells in blood vessels cause a Doppler shift of incident laser light, which correlates with the velocity of blood flow. Continuous Wave Ultrasound involves an ultrasound signal, which is transmitted into the skin. The change in frequency of the reflected signal with respect to the transmitted signal provides an indication of blow flow. The objective of this research was to examine a method for the detection of arterial blood flow, which utilized the 4-electrode sensor for the measurement of Tissue Impedance or the Bio-impedance method. The system developed, for the synchronous detection method, consisted of both analog hardware and software tools. The analog hardware utilized synchronous detection. ^^The software monitored and performed mathematical operations on the retrieved data. The system developed during this research demonstrated the ability to measure the pulsatile impedance of the skin and present the results in a fashion useful to healthcare providers.
502
Thesis (M.S.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 92 pages.
590
Adviser: Wilfrido Moreno, Ph.D.
653
Impedance plethysmography.
Perfusion skin monitoring.
Tissue impedance.
Matlab.
LabView.
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1385