close

Вход

Забыли?

вход по аккаунту

?

Noninvasive blood-glucose monitoring: A microwave-based biosensor development

код для вставкиСкачать
NONINVASIVE BLOOD-GLUCOSE MONITORING: A MICROWAVE-BASED
BIOSENSOR DEVELOPMENT
by
Poornima Shankar
A Thesis Submitted to the Faculty of the
College of Engineering and Computer Science
in Partial Fulfillment of the Requirements for the
Degree of Master of Science
Florida Atlantic University
Boca Raton, Florida
August 2008
UMI Number: 1457858
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 1457858
Copyright 2008 by ProQuest LLC.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 E. Eisenhower Parkway
PO Box 1346
Ann Arbor, Ml 48106-1346
NONINVASIVE BLOOD-GLUCOSE MONITORING: A MICROWAVE-BASED
BIOSENSOR DEVELOPMENT
by
Poornima Shankar
This thesis was prepared under the direction of the candidate's thesis advisor, Dr. P.S.
Neelakanta, Department of Electrical Engineering, and has been approved by the
members of her supervisory committee. It was submitted to the faculty of the College of
Engineering and Computer Science and was accepted in partial fulfillment of the
requirements for the Degree of Master of Science.
SUPERVISORY COMMITTEE:
^cxJtwc^
Chairperson, Thesis Advisor
^
-chairperson
Member
of
/
/
Electrical Engineering
yLdAsv*^jx&1'
Dean, College of Engineering and
Computer Science
College
Dean, Gradate College
Date
n
/
ACKNOWLEDGEMENTS
I wish to express sincere appreciation to Dr. P.S. Neelakanta for his assistance in
suggesting the research topic and supervising the research work. This effort would not
have been made possible without his guidance and support, in addition to his familiarity
with the needs and ideas on the topic throughout this undertaking. I would also like to
extend my thanks to the committee members, Dr. Salvatore Morgera and Dr. Zvi Roth for
their assistance.
111
ABSTRACT
Author:
Poornima Shankar
Title:
NONINVASIVE BLOOD-GLUCOSE MONITORING: A
MICROWAVE-BASED BIOSENSOR DEVELOPMENT
Institution:
Florida Atlantic University
Thesis Advisor:
Dr. P.S. Neelakanta
Degree:
Master of Science
Year:
2008
This research refers to a proof-of-concept study concerning the development of a
noninvasive blood-glucose monitoring system. The biosensor being considered is a
microwave-based transducer (that can be rendered compatible for ISM band of 2450
MHz and hence Zigbee™ and/or Bluetooth™ compliant). The goal of this study is
tailored to develop eventually a unit for home-based healthcare and/or personalized
wellness monitoring of diabetic patients. This pilot effort is expected to culminate in
future in a wireless hyper/hypoglycemic risk-alert system and possible automatic insulin
infusion pump activation efforts.
The thesis addressed thereof provides details on the fundamentals of sensing
glucose content noninvasively across a finger. The underlying principle of biosensing
refers to detecting the change in the dielectric property of blood with differential changes
in the glucose influx in the finger by sensing microwave (such as 2450 MHz) absorption
and/or reflection so as to assay the glucose content of interest. Preliminary experimental
and theoretical results are presented and discussed.
IV
TABLE OF CONTENTS
TABLE OF FIGURES
VIII
LIST OF TABLES
XI
CHAPTER 1
1
INTRODUCTION
1.1
General
1.2
Motivation
1.3
Approach
1.4
Conceived Biosensing Method
1.5
Objectives of the Research
1.6
Contributions
1.7
Thesis layout
1.8
Cbsure
1
1
2
3
4
5
6
7
8
CHAPTER II
9
BLOOD-GLUCOSE MONITORING: STATE OF THE ART
2.1
Introduction
2.1.1 The Blood and Its Composition
2.1.2 Structure of Glucose
2.1.3 Role of Glucose in Human Body
2.1.4 Diabetic Condition
2.1.5 Blood Glucose Level: Normal and Diabetic Conditions
2.2
Diagnosis of Diabetes
2.2.1 Fasting Plasma Glucose Test (FPG)
2.2.2 Glucose Tolerance Test
2.2.3 Random Glucose test
2.3
Blood-Glucose Monitoring
2.4
Self Blood-Glucose Monitors
2.4.1 Principle of Operation of a Blood-Glucose Meter: Invasive Versions
2.5
Noninvasive Blood-Glucose Monitoring
2.6
Closure
CHAPTER III
9
9
9
12
14
16
17
19
19
20
21
21
22
23
24
25
26
DIELECTRIC RESPONSE OF HUMAN BLOOD WITH CHANGING LEVELS OF
GLUCOSE CONCENTRATION
26
3.1
Introduction
26
v
3.2
Dielectric Permittivity: A Review
3.3
Human Blood as a Dielectric
3.4
Complex Permittivity of Human-blood as a Function of Glucose
Concentration
3.5
Concluding Remarks
CHAPTER IV
27
28
29
31
33
ANATOMICAL DESCRIPTIONS OF WORK-PARTS (FINGER AND WRIST) IN
THE HUMAN BODY COMPATIBLE FOR NONINVASIVE GLUCOSE
MONITORING
33
4.1
Introduction
33
4.2
Anatomical Description of a Finger
34
4.3
Monitoring Blood-Glucose at the Finger
37
4.4
Approximate Dimensions of an Adult Human Finger
37
4.5
Human Wrist Anatomy
38
4.6
Concluding Remarks
39
CHAPTER V
40
MULTILAYERED, LOSSY-DIELECTRIC CYLINDER REPRESENTATION OF
HUMAN FINGER AND THE HUMAN WRIST
40
5.1
Introduction
40
5.2
Evaluation of EM Absorption Characteristics of Human Finger and/or Wrist.
40
5.3
Inferential Remarks
45
5.4
Concluding Note
45
CHAPTER VI
47
EXPERIMENTAL STUDY AND RESULTS
47
6.1
Introduction
47
6.2
Experimental Efforts of the Present Study
48
6.3
Biomimetic Blood: Design and Fabrication
49
6.4
Biomimetic Structure of Human Finger: Design and Fabrication
50
6.5
Microwave Transmission-line Plumbing to Measure Glucose-based Power
Absorption in the Biomimetic Phantom
52
6.5.1 Test Procedure
55
6.6
Noninvasive Microwave Transmit-receive Sensor Placed on the Biomimetic
Finger: Design Details and Measurements
65
6.6.1 Patch Antenna Assembly Details
68
6.6.2 Patch Antenna Details
71
VI
6.7
Discussions on the Results
6.7.1 General Observations
6.7.2 Theoretical Results
6.8
Inferential Remarks
6.9
Concluding Note
75
75
76
77
77
CHAPTER VII
79
WHAT LIES AHEAD?
7.1
Introduction
7.2
Underlying Hypotheses of the Present Study
7.2.1 Glucose Content Alters Dielectric Property
7.2.2 Critical Questions
7.3
Any Percentage Change in Blood Constituents can be Sensed via EM
Absorption Measurements
7.3.1 Critical Questions
7.4
Directions for Future Study
7.5
Closure
CHAPTER VIII
79
79
79
80
80
81
81
82
83
84
EXECUTIVE SUMMARY
84
REFERENCES
87
Vll
TABLE OF FIGURES
Figure number
Page
Figure 2.1(a): Fisher projection of the chain form of D-glucose
13
Figure 2.1(b): Chain form of D-glucose
13
Figure 2.2(a): a-D-glucopyranose
13
Figure 2.2(b): /3-D-glucopyranose
13
Figure 2.3: The role of insulin
15
Figure 2.4: Commercial version of a finger-prick type blood-glucose monitor
22
Figure 4.1: Human Finger
34
Figure 4.2: X-Ray of human palm
35
Figure 4.3: Vascular system of the palm
36
Figure 4.4: Detailed structure of a finger
36
Figure 4.5: Conceived thimble-housed noninvasive sensor arrangement on the finger...37
Figure 4.6: Approximate dimensions of an adult finger
38
Figure 4.7(a): Approximate dimensions of the wrist region
39
Figure 4.7(b): Anatomical description of human wrist
39
Figure 5.1: EM excitation of the work part
41
viii
Figure 5.2: Two dimensional representation of the work part
43
Figure 6.1(a): Biomimetic model of human finger constructed
51
Figure 6.1(b): A photograph of the constructed biomimetic model of human finger
52
Figure 6.2(a): Microwave (X-band) plumbing
53
Figure 6.2(b): Photograph of the microwave transmission plumbing arrangement
54
Figure 6.2(c): Experimental set up
54
Figure 6.3: Housing unit used to place the biomimetic finger in the experiment
55
Figure 6.4(a): Graph showing the variation of microwave absorption with changing
glucose concentration in two sets of trials performed
58
Figure 6.4(b): Normalized microwave absorption as observed in the two experimental
trials
59
Figure 6.4(c): Plots showing the observed normalized microwave absorption in terms of
relative microwave power in dB
59
Figure 6.5(a): Plots of microwave absorption with changing glucose concentration in
two trials
61
Figure 6.5(b): Plots of normalized microwave absorption with changing glucose
concentration in two trials
61
Figure 6.5(c): Plots of normalized microwave absorption in dB with changing glucose
concentration in two trials
62
Figure 6.6: Plot of the average values of the readings obtained in the four trials
elucidated via logarithmic clustering
64
Figure 6.7(a): Patch antenna assembly
66
Figure 6.7(b): A cross sectional view of the antenna structure covered by an
enclosure
67
IX
Figure 6.8: Photograph of the phantom housed in the thimble-like enclosure
68
Figure 6.9(a): Patch antenna
70
Figure 6.9(b): Photographs of the other antennas tested
70
Figure 6.10(a): Microwave absorption versus glucose concentration in saline solution
using patch antenna arrangement
74
Figure 6.10(b): Microwave absorption versus glucose concentration in water using
patch antenna arrangement
74
Figure 6.11: experimental set up with patch antenna based thimble arrangement
75
x
LIST OF TABLES
Table number
Table 2.1: Composition of blood
Page
11
Table 3.1: Composition of blood
28
Table 5.1: Electrical characteristics of biological parts
44
Table 6.1: Values of permittivity and conductivity at the respective frequencies
according to the standards of EN 50361 and IEEE 1528-200x
49
Table 6.2: Results obtained using the experimental set up of Figure 6.2(c)
57
Table 6.3: Set of experimental results obtained using a single biomimetic
finger
60
Table 6.4: Normalized readings obtained from the tables 6.2 and 6.3 to plot Figure
6.6
63
Table 6.5: Measurements using the patch antenna with varying amounts of glucose in
saline solution
72
Table 6.6: Measurements using the patch antenna with varying amounts of glucose in
water
73
XI
CHAPTER I
INTRODUCTION
1.1
General
Contemporary blood-glucose monitoring towards diabetes management as a personalized
healthcare regimen is highly desirable for use in household premises as well as in care
centers for aged and disabled. Diabetes (Types I and II) has an enormous impact on the
lives of both juveniles and adults and needs proactive personalized/individualized
wellness support. This is highly warranted in situations when patients are not awake and
in children and/or disabled/aged persons who inadvertently or otherwise not feeling (or
becoming aware of) hypo-/hyper-glycemic state, that may lead them to get into a lifethreatening stage. Glucose monitoring is also extremely important in patients in intensive
care units and those being fed with intravenous glucose drips. It has also been reported
that hypo or hyperglycemic state may prevail in HIV/AIDS patients due to medications
or otherwise.
Tight control of glucose is much needed in all the aforesaid situations; and, blood glucose
monitoring is currently being done by taking frequent blood samples from patients to
measure the glucose levels in them. Relevant procedure is both uncomfortable for the
patients, tediously time-consuming for the healthcare staff and expensive for hospitals
and clinics. While alert patients can test themselves with simple, finger-prick glucose
1
monitors, the others mentioned above would need a noninvasive method that
automatically assesses the glucose level (intermittently or continuously) without any
intervention of the self and/or patient-care personnel. Such a unit can also offer an
alerting annunciation; and, if needed, it can further activate a patient-worn insulin
infusion pump.
1.2
Motivation
Inspired by the need indicated above, this research is conceived to address the proof-ofconcept efforts towards developing eventually a microwave-based, noninvasive biosensor
unit that can continuously monitor the blood sugar in a subject; and, the device could also
be tailored if necessary to communicate wirelessly, the sensed information to an indoor
base-station for monitoring, offer alert annunciation and if needed (wirelessly) activate
the insulin infusion pump on the patient. The concept of this microwave-based design can
be ultimately rendered to be ZigBee™ or Bluetooth™ compliant. (The ZigBee™ and
Bluetooth™ are license-free, ISM band (2.45 GHz) wireless transceivers [1.1]). In the
perspectives of today's technology, the aforesaid systems are based on CMOS VLSI and
can be adopted to include a (bio) sensor so as to monitor the blood chemistry in vivo and
hence sense the glucose level in blood. As a conceived objective, it is proposed that, this
entire microwave-based (and eventual ZigBee™ or Bluetooth™ compliant unit) be
placed as a "thimble" on the patient's finger or wrist-strapped. An antenna on the finger
surface (or on the wrist) will send EM energy (such as 2.45 GHz) signal into the finger
(or the wrist) and will assess the chemistry of the blood-stream; that is, the signature of
the absorbed and/or reflected signal will be assayed for the extent of glucose content.
2
This assay can be performed at a remote (indoor) monitor/base-station wirelessly
connected to the biosensor (for example via ZigBee™ or Bluetooth™ protocols).
The U.S. Food and Drug Administration (USFDA) has approved several forms of noninvasive blood glucose monitors, among which "applying radio waves" is one of them
[1.2]. The present concept belongs to this category. It is conceived to be totally
noninvasive and uses the microwave spectrum of RF/EM energy to measure the glucose
concentration in blood. It is also a novel strategy and differs from other noninvasive
glucose-monitoring efforts reported in the literature [1.3] that use RF impedance
spectroscopy at lower side frequency band (kHz to MHz). That is, the proposed method,
by using a microwave technique assesses the RF transmission and/or power-loss
characteristics of the blood. This electromagnetic power absorption leads to eventual
measurement of the extent of glucose content present in the blood. Thus it differs from
RF impedance spectroscopic method of [1.3].
1.3
Approach
The research as reported in this thesis refers to a first- level effort in which the relevant
biosensor is conceived with appropriate microwave instrumentation. The microwave
based technique has eventual feasibility of using a ZigBee™/Bluetooth™ based device.
Testing is attempted in this with non-biological biomimetic sample of blood concocted
with a recipe made of polyethylene glycol (PEG), saline solution, mineral oil and glucose
solution. No human/animal subjects are used. It is suggested that a further effort can be
taken up as a follow-up study wherein a system of cell-culture can be used to emulate the
blood; and a prototype can be devised for clinical trials with the collaboration of
3
appropriate bio medical/medical personnel/institutions (such as University of Miami
School of Medicine, Diabetes Research Institute).
1.4
Conceived Biosensing Method
As mentioned before, this research is undertaken to develop a biosensor that can sense
the differential blood-glucose concentration across a finger (or at the wrist) noninvasively. The underlying principle hypothesized is based on the fact that the dielectric
property of the human-blood would change vis-a-vis the glucose content. The humanblood can be modeled as a composite dielectric made of solid contents such as red blood
corpuscles (RBCs), erythrocytes, white blood corpuscles (WBCs), platelets, inorganic
solutes (such as sodium, potassium) and organic solutes (such as lymph) as described by
Neelakanta et al. in [1.4]. Other constituents remaining invariant, assuming that the
glucose content changes, it can be expected, that the effective complex permittivity of
blood would change. Therefore, it is surmised that any electromagnetic (EM)
transmission or reflection across the blood would indicate a discernable artifact when
detected.
It is envisaged in the present study to use a compatible 2450 MHz ISM-band microwave
source for the intended EM interaction with the complex permittivity characteristic of the
blood. For example, the ISM-band sources(s) can be chosen, for the reasons that it is a
permitted frequency for medical applications and opens up the possibility of using
Zigbee™/Bluetooth™ chips that are commercially available for the implementation in
the project.
4
1.5
Objectives of the Research
Commensurate with the motivation of the research indicated above, the objectives carried
out and described in this thesis are as follows:
•
Enunciation of the research concept: For monitoring blood-glucose level
autonomously (without the intervention of the patient or supervisory medical
staff) a noninvasive approach is imminent. As mentioned earlier, this is more so
important for patients who are asleep and/or disabled. The invasive and partiallyinvasive methods in vogue are of essentially finger-prick type. The infra-red
systems that prevail are quasi-invasive. On the contrary, the concept indicated
here conforms to be fully noninvasive in biosensing the glucose content in the
blood of a subject. The associated strategy when pursued in the microwave ISM
band will also facilitate possible wireless monitoring and control of glucose level
in the patient
•
Analytical considerations: As stated before, the biosensing concept being studied
refers to assessing the dielectric characterization of the human-blood with varying
amount of glucose content. Relevant electromagnetic considerations are modeled
with reference to a multi-layer representation of, for example, the human finger
accommodating the biosensor. The analytical formulation derived would lead to
the design strategies of the system
•
Design of a proof-of-concept unit of the test biosensor and related electronics:
This task forms the objective as the first phase of research leading to eventual
prototype development of the sensor
•
Construction of a proof-of-concept unit and testing: Concurrent effort conforms to
making a unit and testing it towards elucidating the feasibility of the concept
•
Inferential conclusion: The theoretical and experimental efforts are compiled and
the closure of the thesis is written to indicate the positive and the negative aspects
of the pursued efforts, suggest refinements to the concept-design and enumerate
tasks for further studies
5
•
Summary of research tasks: The primary research tasks addressed in this work
includes the following:
o
Concept and evaluation of a proof-of-concept test-sensor at X-band
frequencies
o
Design of compatible instrumentation to assess the monitored signal
o
Simulation of a glucose dosage influx in a test-tube model of the finger
and detection of the microwave-based signal through the test-tube
o
Modeling human-blood with varying extents of glucose content
o Modeling a layered structure of human finger or wrist for EM interception
o
Suggestions to translate the underlying biosensing at
ISM band for
possible use with ZigBee™ or Bluetooth™ source
o Analysis of the data gathered and inference.
(Note: Though the present study, as mentioned before, can be extended to a wrist
strapped strategy, the study performed is confined only to finger-thimble model. The
wrist strap version is indicated for a future study).
1.6
Contributions
As stated earlier this is a preliminary and proof-of-concept study carried out to develop a
totally noninvasive blood-glucose monitor using a microwave-based biosensing
arrangement. Consistent with this objective, the study performed has led to the following
contributions:
•
Primarily it is identified from open literature and by experiments performed that
blood glucose content could change the dielectric properties of the blood.
Statistical mixture model developed here confirms this observation
•
By representing a compatible anatomical part such as the finger as a multilayer
dielectric structure (in which blood is a constituent along with tissues, bones etc.),
it is shown that a fractional change in the complex dielectric property of the
6
structure would cause corresponding change in electromagnetic absorption
characteristics. This modeling enables an inference that with change in glucose
content in the anatomical part, a corresponding electromagnetic loss can be
expected
•
While the existing glucose sensing devices are invasive, the present study
conceives a noninvasive method for the same purpose. This contribution refers to
a noninvasive sensing of EM absorption by the finger versus varying glucose
content
•
The noninvasive approach indicated is tested with a biomimetic model of finger
made of blood imitating phantoms
•
The noninvasive electromagnetic sensor developed works at microwave
frequencies (compatible with ISM bands). Relevant thimble-like device housing
the EM transceiver is designed, tested and evaluated
•
Though the present study is of preliminary nature, it offers directions for further
work. Also indicated are the precautions and pitfalls as well as suggested design
changes in future work.
1.7
Thesis layout
This thesis is written in eight chapters with the contents as listed below:
Chapter I
Introduction: This outlines the scope of the work and summarizes the
motivation, approach and contributions of the research work.
Chapter II
Blood-glucose monitoring: State of the Art: In this chapter,
presented are details on blood-glucose chemistry and the principle
behind blood-glucose monitoring.
Chapter III
Dielectric Response of Human Blood with Changing Levels of
Glucose Concentration: Addressed here are theoretical aspects of
dielectric permittivity of a test substance, such as blood, in response to
changing levels of constituent content, like glucose.
7
Chapter IV
Anatomical Descriptions of Work-parts (Finger and Wrist) in the
Human Body Compatible for Noninvasive Glucose Monitoring:
Since the biosensing of glucose has to be done in vivo, the compatible
anatomical parts are identified in this chapter. They refer to the finger,
wrist and/or ankle. The basic anatomical details as needed are
presented.
Chapter V
Multilayered, Lossy-dielectric Cylinder Representation of the
Human Finger and the Human wrist: In this chapter, a multilayer
model of the anatomical part of interest is presented and the EM
absorption characteristic of the model is elucidated.
Chapter VI
Experimental Study and Results: This chapter is written in detail to
present the experimental studies performed, test results gathered and
inferences made.
Chapter VII
What Lies Ahead?: This chapter identifies open questions, discusses
critical aspects of the study and gives directions for future work
Chapter VIII Executive Summary: This executive summary is a gist that
summarizes the content of the thesis.
1.8
Closure
This (Chapter I) is written to offer a general overview on the scope of the research and
the underlying objectives. It also outlines the contents that are deliberated in the chapters
that follow.
8
CHAPTER II
BLOOD-GLUCOSE MONITORING: STATE OF THE ART
2.1
Introduction
In developing a reliable blood-glucose monitoring system, it is important to study the
basic characteristics of blood, its composition and its role in human body. Assaying the
blood-glucose in vivo, being the core theme of the present study, the chapter is written to
outline the blood glucose chemistry and its characteristics. Further discussed are details
pertinent to the blood glucose-monitors presently available in the market, listing their
merits and demerits.
2.1.1
The Blood and Its Composition
The main functions of blood in a living system are transportation of nutrients, waste
products, gases and hormones, regulation of body temperature, maintenance of acid-base
balance and protection against pathogens. An adult human has about 6 to 8 liters of blood
of which 38% to 48% is composed of blood cells and the remaining is blood plasma. The
liquid portion of blood is approximately 91% water.
Blood is more viscous than water due to its cellular content and proteins present in the
blood plasma. It is also slightly alkaline with a pH of 7.35 to 7.45.
9
•
Cellular content in blood: As mentioned before, the cellular content make up 38%
to 48% of blood. They are composed of
1. Red blood cells (RBCs), also known as erythrocytes are in the shape of
biconcave discs. A normal RBC count ranges from 4.5 to 6.0 million cells
per mm3. Hematocrit value is a measure depicting the amount of RBCs in
blood. It is determined by a single test that involves drawing a small
amount of blood and centrifuging it to force the cells to separate from the
blood plasma, thus enabling to determine the percentages of cells and
blood plasma. A healthy hematocrit range is 38% to 48% inasmuch as
RBCs make up most of the cellular content in the blood. Red blood cells
contain hemoglobin, which is a protein that contains iron, enabling them
to transport oxygen
2. White blood cells (WBCs), also known as leukocytes play a major role in
protecting the body against pathogens. There are five kinds of WBCs and
a normal WBC count ranges from 5000 to 10,000 cells per mm3. The
different kinds of leukocytes are neutrophils, eosinophils, basophils,
lymphocytes and monocytes
3. Platelets also known as thrombocytes are cell fragments that play a major
role in preventing blood loss when a vein or artery is severed or when a
capillary ruptures.
•
Blood plasma: Plasma is the liquid portion of blood. It is made up of 91% water
and transports nutrients such as glucose, waste products such as dissolved carbon
dioxide in the form of HCC>3~ ions, hormones etc. It also contains plasma proteins
such as prothrombin and fibrinogen which aid in blood-clotting, albumin that
maintains osmotic pressure of blood, globulins that carry fat to the tissues and
antibodies secreted by the WBCs. Blood plasma is also responsible for
maintaining body temperature.
10
The blood composition is tabulated in Table 2.1 [2.1].
Table 2.1: Composition of blood [2.1]
Contents
Cellular elements
Erythrocytes (RBC)
Leukocytes (WBC)
Monocytes
Granulocytes
Lymphocytes
Concentration
5 x 106 cells/mm3
5 to 8 * 103 cells/mm3
2.5 to 5 x 105
cells/mm
•5
Platelets
Plasma Molecular
constituents
Albumin
Globulin
Lipoprotein
Fibrinogen
Glucose
3.5-5.3 mg per 100ml
2.1-3.3 mg per 100ml
0.2-0.4 mg per 100 ml
70-120 mg per 100ml
0.9%
Ionic Content of plasma
Na+
K+
Ca2+
cr
HCCV
HPO 4 " 2
other cations
other anions
11
2.1.2
Structure of Glucose
Glucose is a simple sugar also called as monosaccharide having six carbon atoms. It is
the chief energy source for all living cells in the body. The chemical formula of glucose
is C6H12O6. The different structures of glucose are shown in Figures 2.1 and 2.2. (The socalled Fischer projection shown in Figure 2.1(a) is a method of representing a threedimensional structure on a two-dimensional plane by projection). Since glucose contains
the aldehyde group (CH=0), it is also called as an aldose. The different ring structures
taken up by the glucose molecules are due to the presence of asymmetric carbon atoms.
Glucose possesses four asymmetric carbon atoms known as chiral carbons. This results in
glucose having sixteen different structures known as stereoisomers [2.2]. In biological
systems, a particular isomer namely, D-form of glucose predominates.
The straight chain structure of glucose is observed only in dry powder form. In aqueous
solution glucose takes the ring structure also depicted as cyclic chair form as shown in
Figure 2.2. The cyclic form of glucose is also called as glucopyranose since the ring
contains 5 carbon atoms (the 6th carbon atom being outside the ring). Formation of the
ring results in the creation of an additional asymmetric carbon (C-l) in the molecule and
hence resulting in the isomers a-D-glucose and P-D-glucose. (Same is true for L-Glucose
also, but the structures of L-glucose are not of concern because L-glucose is not found in
biological systems). The new asymmetric center is called the anomeric carbon and the Gland p-isomers are called as anomers. The anomers of glucose are shown in Figure 2.2.
The a and p forms of glucose are inter-convertible as a spontaneous phenomenon is
called as mutarotation. With respect to D-glucose, either anomer, when dissolved in
12
water will slowly undergo mutarotation until an equilibrium mixture consisting of one
third a-D-glucose and two thirds [3-D-glucose is attained.
0_.H
-OH
HO-
OH
-H
H-
-OH
H-
-OH
OH
OH
O
OH
v
OH
Figure 2.1(a): Fisher projection of
the chain form of D-glucose [2.3]
Figure: 2.1(b): Chain form of D-glucose [2.3]
Figure 2.2(a): a-D-Glucopyranose [2.3]
Figure 2.2 (b): (3-D- Glucopyranose [2.3]
For the purpose of the present research, anhydrous a-D-glucose 96% procured from
Sigma-Aldrich Inc has been used.
13
2.1.3
Role of Glucose in Human Body
Glucose is the chief energy source for all cells in human body. In fact, human body
breaks down polysaccharides, starch and other carbohydrates into glucose before their
usage in cells and tissues. The digestion of starch and carbohydrates begin in the mouth
with the mixing of the enzymes in saliva (amylase) and the starchy food. It works to
break down starch into disaccharides such as maltose and sucrose. Further digestion in
stomach, small and large intestine breaks down the starch into glucose.
Glucose is then directly absorbed into the blood stream. The concentration of glucose in
the blood stream is closely regulated to fall in the range of 70-140 mg/dl. (The
concentration of glucose is either expressed in mg/dl or in mmol/L. Conversion from
mg/dl to mmol/L can be done by dividing by 18. Throughout this thesis blood-glucose
concentration is expressed in mg/dl).
The blood-glucose level is not a constant; for example, just after a meal, the bloodglucose value would increase, sometimes close to 140 mg/dl and then may fall down
close to 70 mg/dl in 8 hours after a meal. Immediately after a meal, when the glucose
enters the blood stream, the rise in the concentration of glucose in blood triggers the beta
cells of the pancreas to secrete the hormone insulin into the blood. Certain cells in the
body require insulin to bind onto the insulin receptors of the cells in order to facilitate the
entry of glucose into the cell from the blood stream. This is pictorially described in the
Figure 2.3 [2.4]
14
Figure 2.3: The role of insulin [2.4]
Also influenced by the insulin is the storage of glucose as a polysaccharide glycogen by
the liver and muscle cells. Hence the presence of insulin lowers the amount of glucose
concentration in the blood.
When glucose concentration reduces beyond a limit of 70 mg/dl in the blood, another
hormone secreted by the alpha cells of the pancreas namely, the glucagon catalyzes the
conversion of glycogen into glucose in the liver and thus brings about a raise in blood
glucose concentration. Other hormones that raise the blood glucose concentration are
somatostatin, epinephrine, Cortisol, adrenocorticotropic hormone and growth hormone
[2.5]. All these hormones collectively work to maintain an optimal level of glucose
concentration in the body.
15
2.1.4
Diabetic Condition
The term diabetes mellitus is used to describe a disease in which the body fails to control
the blood glucose concentration due to insufficient insulin or due to the cells becoming
less responsive to insulin. As mentioned in Chapter 1, there are two types of diabetes:
Type I diabetes results when there is insufficient or no production of the insulin hormone
by the beta cells of the pancreas. Lack of insulin in blood results in the cells that are
dependent on insulin for absorption of glucose from blood, getting depleted of their
energy source, and further, it also results in the increase in blood glucose levels. In order
to lower the concentration of glucose in blood, the kidneys remove glucose from the
blood along with water, thus dehydrating the body. The exact causes for Type I diabetes
are unknown. It is also known as juvenile diabetes since its occurrence has been observed
in young people (It may also occur in older people). Treatment of this type of diabetes is
done by supplying insulin to the body externally (using insulin pumps, injection etc).
Type II diabetes results when the cells that depend on insulin for absorption of glucose
from the blood, become partially resistant or completely resistant to the hormone. In
other words, insulin is no longer able to bind on the insulin receptors of the cells thus
increasing the blood glucose levels at the same time depleting energy source for these
cells. The pancreatic cells, when faced with increase in the blood glucose concentration
start producing more insulin thus trying to compensate for the rise. When faced with this
condition for longer periods of time, the beta cells of the pancreas might slowly get
destroyed resulting ultimately in Type I diabetes. The causes for Type II diabetes are
16
mainly attributed to genetic factors and life style. This diabetes can however, be
controlled by careful diet, exercise and with medication to lower the insulin resistance.
Gestational diabetes is seen in pregnant women having relatively weak pancreas. This
diabetes signals that the person may get Type I or II diabetes in the future and hence the
patient should change her lifestyle in order to prevent it from happening. The blood sugar
levels generally return to normalcy with the completion of the pregnancy; but it should be
carefully monitored and maintained during gestation inasmuch as there is a higher risk of
miscarriages among individuals having gestational diabetes.
The symptoms of diabetes are as follows [2.6]:
1. Frequent and excessive urination
2. Excessive intake of water and perpetual thirst
3. Weight loss
4. Fatigue
5. Numbness of hands or legs
6. Blurry vision
7. Frequent cuts and bruises that take long time to heal
8. Dry and itchy skin.
2.1.5
Blood Glucose Level: Normal and Diabetic Conditions
The normal concentration of glucose in the blood ranges from 70 mg/dl to 140 mg/dl.
When the blood-glucose concentration of a person falls below 60 mg/dl, the person is
said to be in the hypoglycemic stage. As mentioned before, glucose is the primary source
of energy in the body. The liver and the muscle cells store excess glucose in the form of
glycogen as a short term energy reserve. When the blood sugar level drops, liver converts
17
glycogen into glucose, which is further absorbed by the blood and hence raises the blood
sugar level. When the blood sugar levels continue to drop, the brain and the central
nervous system are depleted of their energy source. The primary energy source for the
central nervous system is the glucose present in the blood. They do not have any store of
glycogen like the muscle cells. Hence the brain signals the release of other hormones that
raise blood-glucose concentration such as Cortisol, epinephrine etc. When the blood sugar
level is close to 60 mg/dl, the person may experience excessive sweating, palpitation,
nervousness, hunger, difficulty in speech etc. all of which are a signal from the brain that
the person should eat food and increase the blood glucose levels. If this warning sign is
ignored and the blood-glucose levels continue to drop, then the person may get drowsy,
confused and may even go into coma due to the impairment in the function of the brain
and the central nervous system. Hypoglycemia may result in patients with Type I or II
diabetes whose diet and medication are not adjusted properly.
When the blood-glucose level is in excess of 150 mg/dl, the person is said to be in the
hyperglycemic state. Extended periods in this state result in excessive urination and
hence excessive thirst (because of kidneys that try to lower the concentration by way of
urination), headaches, itchy skin, blurry vision, cuts and bruises taking a long time to
heal, feeling of weakness and fatigue etc. This condition is prevented by diet and
exercise. If the symptoms are due to the onset of diabetes, then the person needs to take
appropriate medication since neglecting these symptoms would lead to further
complications such as destruction of beta cells of the pancreas and ketoacidosis, (a
condition where fats begin to get converted into glucose along with the production of
18
ketones). Hyperglycemia may also result due to the destruction of the beta pancreatic
cells in which case the person is said to have Type I diabetes and may need supply of
insulin to the body externally.
Thus, it is very important to maintain the blood-glucose concentration at the normal range
(70 mg/dl to 140 mg/dl) for healthy functioning of the body and mind.
2.2
Diagnosis of Diabetes
Early diagnosis is the first step towards treating diabetes. The National Institute of
Diabetes and Digestive and Kidney Diseases advises that anyone who is 45 years or
older, and has risk factors such as overweight, have a parent, and/or siblings with
diabetes, low extents of exercise, high blood pressure, had prior gestational diabetes
should consistently test for diabetes. Also, anyone who is younger and who has the
symptoms such as frequent urination, perpetual thirst, too much sweating, fatigue,
unexplainable loss of weight, very dry skin and bruises that take very long to heal must
also get tested for diabetes. The various tests that are done to diagnose diabetes are as
follows:
2.2.1
Fasting Plasma Glucose Test (FPG)
This test involves taking a sample of blood from the vein of an individual who has fasted
for more than 8 hours (done preferable in the morning), and measuring the concentration
of glucose in the blood. If the blood-glucose level is more than 126 mg/dL, then the
person has diabetes. A blood-glucose level of 100-125 mg/dL is pre-diabetic meaning
that the person has a very high risk of developing Type II diabetes and needs to alter
19
his/her lifestyle as well as food habits and regular exercise in order to bring down the
blood-glucose level. A blood-glucose level of 99 mg/dL and less of FPG is normal.
The positive result for fasting plasma glucose test for diabetes is confirmed by
performing the same test on another day. FPG is the preferred test by patients due to the
convenience of performing the test.
2.2.2
Glucose Tolerance Test
It is a test involving administration of glucose (orally) and measuring how quickly it is
cleared from the blood. Since the administration of glucose is done orally, it is also called
as Oral Glucose Tolerance Test or OGTT.
The procedure for the test is to have the patient fasting for 8 to 14 hours (Like the Fasting
Plasma Glucose test, this test is also done in the morning and involves drawing a blood
sample of the patient). This sample is labeled as the zero-hour sample and its glucose
level is measured. The patient is then given a glucose solution to drink. The solution
consists of 1.75 g of glucose/ kg weight of the patient up to a maximum of 75 g of
glucose. The test requires the patient to consume the solution within 5 minutes. After two
hours, a blood sample of the patient is drawn, labeled as the two-hour sample and its
blood-glucose level is noted.
Blood-glucose levels of more than 126 mg/dl in the zero-hour sample and more than 200
mg/dl in the two-hour sample confirms diabetes. Normal levels are less than 110 mg/dl in
the zero-hour sample and less than 140 mg/dl in the two-hour sample. Patients having a
blood-glucose concentration between 110 mg/dl and 126 mg/dl in the zero-hour sample
and between 140 mg/dl and 200 mg/dl in the two-hour sample are said to have impaired
20
glucose tolerance. These patients are in the pre-diabetic stage and are at a high risk of
developing diabetes.
2.2.3
Random Glucose test
Random glucose test is performed any time during the day. The blood sample is drawn
and blood-glucose levels are measured at any point of time during the day. If the bloodglucose level is more than 200 mg/dl and the patient has the symptoms of excessive
thirst, urination and weight loss, the patient is suspected to have diabetes. It is then
confirmed by FPG test or OGTT.
2.3
Blood-Glucose Monitoring
In persons with diabetes or in a condition where they are unable to maintain blood
glucose level in the normal range (for example, in patients who are administered glucose
intravenously), blood-glucose levels need to be periodically checked and appropriate
action need to be taken to prevent the risky conditions of hyperglycemia or
hypoglycemia.
Patients with Type I diabetes inject insulin externally and hence are prone to fluctuations
in their glucose levels. They are generally advised to check their blood glucose level
more than three times each day. Patients with Type II diabetes need not monitor as often,
but need careful monitoring whenever medication, diet and lifestyle changes are
implemented. Patients in the intensive care of the hospital, who are administered glucose
intravenously along with other drugs, need to be kept in close monitoring since, in
addition to the risk of development of hyperglycemia, the drugs administered to them
21
intravenously will work effectively only when the blood-glucose level is in the normal
range.
2.4
Self Blood-Glucose Monitors
Patients today have a wide range of choices that help them to monitor their blood-glucose
levels. For self- monitoring of blood-glucose, finger- prick type of blood-glucose monitor
is most commonly used. It contains a test strip on which a small sample of blood obtained
from a finger-prick is placed. The test-strip has coated certain reagents that facilitate the
measurement of blood glucose level such as glucose oxidase, dehydrogenase, hexokinase
etc. These chemicals react with the blood and glucose is measured in terms of the electric
conductivity of the blood sample reacting with the strip chemicals or by the amount of
light reflected from the blood sample.
A typical commercially available blood-glucose monitor is shown in the Figure 2.4 [2.7].
Figure 2.4: Commercial version of a finger-prick type blood-glucose monitor
22
2.4.1
Principle of Operation of a Blood-Glucose Meter: Invasive Versions
Most finger-prick type of glucose meters are electrochemical sensors. That is, the sensor
consists of two potentiometric ion- selective electrodes. Such electrodes that are used in
glucose monitors are sensitive to Oxygen (O2) [2.8]. The concept of an electrochemical
biosensor was first suggested by Clark [2.9]. His method refers to building a biosensor by
using an oxygen sensor along with an enzymatic system. Kuhn further describes in [2.7]
that any glucose monitor would require three elements, namely:
1. Glucose specifier - The test-strip coated with glucose oxidase or glucose
hydrogenase acts as the specifier for glucose monitors. This comprises of the
enzymatic system of the biosensor. The specifier takes care of the requirement
that the measurement would be very specific to the changes in the concentration
of glucose alone, and change in concentration of any other element in blood, for
example sodium ions, potassium ions, urea, hemoglobin etc. will not affect the
measurement
2. Electron shuttle or mediator: In certain glucose meters (such as Accu-chek™)
glucose meters, potassium ferricyanide is used as mediator
3. Stabilizing agents for the glucose specifier: The specifier indicated above consists
of an enzyme that requires optimum conditions of temperature humidity etc to
function as expected. These stabilizing agents ensure a long shelf-life to the
specifier
4. Electrodes: Two identical electrodes are used for amperometric detections. The
amount of current flowing across these two electrodes gives the measure of the
concentration of glucose in the blood sample. The electrodes are made of
materials like palladium.
When a sample of the blood-glucose is placed on the strip, it is absorbed into it due to the
capillary action. (The surface of the strip has a capillary roof in order to ensure this). The
23
glucose in the blood reacts with the enzyme glucose oxidase to produce gluconic acid and
a reduced form of glucose oxidase. The reduced glucose oxidase reacts with potassium
ferricyanide to produce potassium ferrocyanide. The working electrode, which has a
higher positive potential than the rest potential of the mediator, oxidizes the mediator
back into potassium ferricyanide (from potassium ferrocyanide). This gives rise to a
differential current flowing between the electrodes and this current is proportional to the
amount of glucose present in the blood.
The aforesaid method of testing blood-glucose level is essentially an invasive bloodglucose monitoring inasmuch as it involves pricking the finger and extracting some
amount of blood from the subject.
2.5
Noninvasive Blood-Glucose Monitoring
Another kind of blood glucose monitoring approved by US Food and Drug
Administration is noninvasive monitoring of blood glucose level, which does not involve
puncturing the skin to extract blood and measure the blood-glucose level. The advantages
of non-invasive blood glucose monitoring are:
1. It eliminates the pain of finger-pricking
2. It is more hygienic since it does not require a clean needle to prick finger
3. Can be used for continuous blood-glucose monitoring, even while sleeping
4. Since pricking of finger is eliminated, it is autonomous of patient/health-care
personnel intervention and glucose-level monitoring can be done as many times
as needed. Hence management of diabetes is done more effectively.
24
2.6
Closure
This chapter offers underlying basis of blood-glucose monitoring. The scope of interest
for the present research refers to the noninvasive blood-glucose monitoring strategy. The
proposal thereof is based on the suggestion/hypothesis that blood-glucose level can be
sensed noninvasively by an external means of exciting the blood flowing in a human part
(such as finger/wrist) with microwave energy and detecting the dielectric response of the
blood stream and hence the glucose content in the blood. The next chapter offers details
on the approach pursued.
25
CHAPTER III
DIELECTRIC RESPONSE OF HUMAN BLOOD WITH CHANGING LEVELS
OF GLUCOSE CONCENTRATION
3.1
Introduction
Commensurate with the theme of the present study, it is required to verify the
hypothesized consideration that the dielectric response of human blood would change
with the level of glucose concentration. Hence addressed in this chapter is a model that
depicts the dielectric permittivity of human blood versus varying extents of glucose level
normally encountered in the underlying regulatory metabolism directing glucose infused
blood stream to various parts of the body.
The dielectric model of blood envisaged here is based on the work done by Neelakanta et
al. [3.1] wherein the human blood is depicted as a statistical mixture with a matrix
structure formed by the particulate substances, molecular plasma constituents and ionic
contents. Glucose, which is a part of the molecular plasma, is currently considered as a
variable in the statistical mixture and hence a corresponding model is deduced to evaluate
the permittivity characteristics of the human blood using the statistical mixture
considerations indicated in [3.1].
By deducing the dielectric properties of blood versus glucose, relevant results would lead
to evaluating the electromagnetic absorption by the blood (in vivo) as a function of
26
glucose concentration. Sensing such absorption characteristics is being considered in the
present study to devise a compatible non-invasive bio-sensor indicated in the earlier
chapters.
3.2
Dielectric Permittivity: A Review
This parameter indicates the behavior of a class of material (medium) in response to an
electromagnetic field. It is decided by the molecular structure of the medium whether to
"permit" the electric field force exerted by the EM energy to which the medium is
exposed. The absolute permittivity of the medium is denoted as e which is equal to e0er
farad/meter where er denotes the (dimensionless) relative permittivity (also known as
dielectric constant) and e0 is the absolute permittivity of free-space and is equal to,
(1/36TI)
x 10" 9 F/M.
In general, e is complex and is depicted as [3.2]:
*
i
€ =€ - j e
it
(3.1)
where e' = e0er and e" is a loss parameter arising as a result of finite electrical
conductivity (a)of the medium and e" = a/coe0erAdding the subscript r to denote relative complex permittivity,
e
*=er-J
o
—
(3-2)
v G>e 0 e r 7
where as indicated before, a is the electrical conductivity of the medium (in S/M) and co
= 2 T x f (with f being the frequency of EM excitation).
27
3.3
Human Blood as a Dielectric
The complex permittivity characteristics of human-blood are studied in [3.1]. For the
purpose of modeling and analysis, the blood is presumed as a suspension of erythrocytes
(red blood corpuscles or RBCs) in a suspending medium of blood plasma. The
composition of human blood as such is approximately depicted in terms of constituents
listed in Table 3.1, (which is the same as Table 2.1)
Table 3.1: Approximate composition of human blood [3.1]
Contents
Concentration
Cellular elements
Erythrocytes (RBC)
Leukocytes (WBC)
Monocytes
Granulocytes
Lymphocytes
Platelets
5x 106 cells/mm3
5 to 8 x 103 cells/mm3
2.5 to 5 x 105 cells/mm3
Plasma molecular constituents
Albumin
Globulin
Lipoprotein
Fibrinogen
Glucose
3.5-5.3 mg per 100ml
2.1-3.3 mg per 100ml
0.2-0.4 mg per 100 ml
70-120 mg per 100ml
0.9%
Ionic content of plasma
Na+
r
Ca 2 +
cr
HCO 3 HPO4 2
other cations
other anions
28
In terms of the components of the human blood shown in Table 3.1, Neelakanta et al.
[3.1] developed a comprehensive model of total blood using statistical mixture theory.
The study addressed in [3.1] thereof is directed at evaluating the dielectric properties of
blood as a function of RBC (erythrocyte content) representing the hematocrit value.
In the present study, however, the model is revised keeping the hematocrit value constant
but changing the glucose content in the blood plasma. Again statistical mixture theory
(originally due to Lichtenecker and Rother [3.3]) is resorted to in developing the model in
question.
3.4
Complex Permittivity
of Human-blood
as
a Function
of
Glucose
Concentration
Referring to Table 3.1, human blood is a composite material depicting the set : {Blood
suspension (RBC/WBC), suspending solution of plasma (organic molecules) and
inorganic (ionic) contents.
Suppose e s> e p and e i denote respectively the relative complex permittivity of
suspension, plasma and ionic contents. Then,
cl=c's-}£s"
(3-3)
ep=e'p-jV
(3-4>
^=€'i-j£i"
(3-5)
29
Further, by considering blood as a composition made of the solid suspensions of RBC
and WBC dispersed in plasma medium, the following approximate mixture formulation
(due to [3.3]) can be specified
t*B=(e*sf{4}~h
(3-6)
where h is the hematocrit fraction (neglecting WBC fraction in relation to dominant RBC
content), and eB is the complex permittivity of the total blood and which can be
identically set equal to (eB' -j €B"). Therefore,
it
T
e
e
n ]h r i
r i
e
e
e
ti
\-h
f
( 3 - 7a )
B - J B = [ S - J s J [ P -J pJ
=>log(CB-jCB) = h x [ € s - J € s ] + ( l - h ) x l o g ( 6 p - € , p )
(3.7b)
Suppose it is presumed that the ionic content remains dissolved in the plasma, then it is
essentially a conductive (lossy) component forming an overwhelming lossy constituent.
Therefore, it can be assumed that,
tj*-jV;
(3.8)
and the net complex permittivity can be written (with inclusion of dissolved ionic
components) as: 6*PI »e'p - j(e"p + e"i). Further, (e"P + e"i).= ffpi /COS'P where o>i is the
conductivity of the plasma constituted by ONG and OQ, which depict the losses contributed
by non-glucose part (NG) and glucose content (G) respectively. Explicitly in terms of
Lichtenecker- Rother mixture formula [3.3],
^P/=[^iVG] 1 " g [^G] g
(3.9a)
Or
30
log(ffPI) = ( l - g ) x l o g ( a N G ) + gxlog(ffG)
(3.9b)
where g is the volume fraction of glucose content.
Equating the real and imaginary parts on either side of equation (3.7b), it can be
approximately shown that,
A8 B
A^
(3.10a)
= K,
where 5B and 8g denote the loss angles of blood and glucose content respectively and Ki
is a constant of proportionality. For small fractional changes of loss angles,
approximately,
APR
—B=K
APg
*
(3.10b)
where PB and Pg are EM power losses in the total blood and due to glucose content
respectively and K2 denotes a proportionality constant.
3.5
Concluding Remarks
This chapter describes the general EM power characteristics of human blood as a
function of the blood content(s). Based on an earlier study [3.1] that the lossy
characteristics of blood can be influenced by its contents (such as hematocrit value), the
present study hypothesizes that, similar power absorption variations can be expected if
blood glucose content changes.
Hence, relevant theoretical considerations presented shows a fractional change EM
absorption by the total blood could be proportionately dependent on the fractional
changes in glucose content, as indicated in equation (3.10).
31
This direct proportionality relation as shown in equation (3.10) could be true only for
small changes in glucose content. However, should there be a significant change in the
glucose content, a non-linear relation between EM absorption versus glucose level can be
expected.
In all, the analysis here implies that, the EM power absorption is directly related to the
lossy dielectric parameter of the constituents of blood. The analysis pursued presumes
that the component of interest (namely glucose) is overwhelmingly significant or
insignificant (as it happens in hyper and hypo-glycemic conditions) in relation to other
constituents. Hence this overwhelming part is considered to be responsible for any
observed EM absorption parameter that can be experimentally elucidated.
Based on the above heuristic it is surmised that, the sensor under consideration can be
designed so that it essentially measures the EM power absorption of blood versus glucose
content. It is designed to be worn externally on a finger or a limb (such as the wrist or
ankle), so as to sense the blood-glucose noninvasively.
32
CHAPTER IV
ANATOMICAL DESCRIPTIONS OF WORK-PARTS (FINGER AND WRIST) IN
THE HUMAN BODY COMPATIBLE FOR NONINVASIVE GLUCOSE
MONITORING
4.1
Introduction
As described in the earlier chapters, the noninvasive biosensing of blood-glucose
warrants a wearable transceiver that can be housed external to a human anatomical part
conveniently. This transceiver assays the EM absorption by the blood-stream in the
anatomical part; and absorption factor is correlated to the glucose content in the blood
stream.
For the purpose of biosensing, three possible locations can be identified:
i.
Finger
ii.
Wrist region
iii.
Ankle region
For example, the sensor can be housed as a thimble on the finger or as a strap on the wrist
(or on the ankle).
In order to facilitate a compatible biosensor geometry on the anatomical part of interest, it
is necessary to understand the anatomy of the test part. Hence indicated in this chapter are
33
relevant details pertinent to the basic anatomical features of the human finger and wrist
region.
4.2
Anatomical Description of a Finger
As shown in the Figure 4.1, the human finger can be regarded as an embodiment of the
following essential constituents:
Distal
joint
4
it
„ , , , , •„ llA
Middle &S \
phalanxHOi 1
Proximo!
*5i
mterProximaTk | l 1
phalangeal phaianxmtJ 1
joint
1 4 1
Figure 4.1: Human Finger [4.1]
The skeletal system: The finger, also called as digit, has typically three shafts
(except the thumb which has two shafts) or long bones called phalanges made of
compact hollow bone (diaphysis) and the hollow canal called medullar cavity
containing yellow bone marrow, which is mostly adipose (fat) tissue. The end
section of the shaft is epiphyses made of spongy bone covered with a thin layer of
compact bone. Between the phalanges are the hinge joints which permit
movement in one plane. Thus the finger comprises of three bones called the
phalanges connected to each other by two joints. These joints are called the
interphalangeal joints. The joint near the end of the finger is called the distal
interphalangeal joint and the joint in the middle between the distal interphalangeal
34
joint and the main knuckle is called the proximal interphalangeal joint. An X-ray
of a human hand is shown in the Figure 4.2.
Figure 4.2: X-ray of human palm [4.2]
•
Muscular system: The muscular system in a finger comprises of extensor muscles
that allow the finger to bend and straighten out, together with the extensor
tendons. (A tendon is a tough fibrous tissue that connects a muscle to the bone,
the extensor tendons connect the phalanges to the extensor muscles). In addition
to this the finger and the thumb joints are covered by cartilage which is a thick
connective tissue that absorbs shock and provides a smooth motion between the
bones. Cartilage is present wherever two bones meet or move against one another.
•
Vascular system: the blood circulation in the finger is mainly by a set of arteries
surrounding the phalanx bones. There are two main branches of arteries entering
the hand which are ulnar artery and the radial artery. They form the superficial
palmar volar arch and the deep palmar volar arch respectively in the palm of the
hand. The arteries that feed the fingers (or the digital regions) branch out from
these volar arches as shown in the Figure 4.3.
35
Proper palmar
dtfulainrlu
*
Racial aitsry
I
prtmir
tularin
pakrornran
Ptinajpa.
POIIGB mVBrf
RMU
Hifeiy
Figure 4.3: Vascular system of the palm [4.3].
The numerous branches of arteries emerge from the volar arches and surround the
phalanges forming a dense section below the nail body as shown in Figure 4.4. The
systemic veins - radial vein and the ulnar vein join to form the brachial vein which drains
blood from the hand and the digital sections.
FINGER
Cuticle
Nail ?ratm Articular Synovial | . „ J J , V S J „ ,, I-x. digitoram
Nail imrt
L-artttage trk'mbww
.'"' tcitdo
Jinm capwte
. Middle pfciianx
_ Pwxunul
plukm
phalanx
Arteries
Nerves
Septa" Joint cavitv
i-1, ttijwtorwn
•- superficial!.-* U-mlun
(•ibTOlis llitjliil'
(plate)
n . digiliirxim
protumlis tendon
Sj'rwx<«l sheath
9i' rt. tersttofis
Figure 4.4: Detailed structure of a finger [4.4]
36
slicath
4.3
Monitoring Blood-Glucose at the Finger
The glucose in the gust of blood from the arteries just below the nail body is normally
sensed in the finger-prick blood-glucose monitoring instrumentation. Presently, the blood
that circulates around the phalanges is used for monitoring the glucose content
noninvasively. The underlying concept is illustrated in the Figure 4.5.
Thimble
Subcutaneous arterial
blood circulation
Figure 4.5: Conceived thimble-housed noninvasive sensor arrangement on the finger
A thimble arrangement (to be described later) shown in the Figure 4.5 houses the
biosensor that noninvasively senses the blood characteristics.
4.4
Approximate Dimensions of an Adult Human Finger
Typically, the finger anatomy can be represented by a multilayered, lossy dielectric
cylinder. The associated average (approximate) dimensions of an adult finger is shown in
the Figure 4.6 .
37
™"»—«*
•«»-***"*'"*
y^jgi&pmW***'
->
Phalange bone complex
t
/
h
v
Dermis
and
epidermis
Subcutaneous
tissue matrix and
Figure 4.6: Approximate dimensions of an adult human finger [4.5]
Length* (1): 74.5 mm; breadth** (b): 14 mm; depth ** (d): 13 mm and Bone diameter
(w): 4 mm
* Total of proximal, medial and distal sections of the finger.
** Average of the values measured at metacarpophalangeal, proximal
interphalangeal and distal interphalangeal joints
4.5
Human Wrist Anatomy
As indicated in [4.6] and [4.7], the approximate anatomical features of the human wrist
can be depicted and modeled as shown in Figure 4.7.
38
4v
Skin
6.35 cm
8.9 cm
*
~¥
0.25 cm
Figure 4.7(a) Approximate dimensions of the wrist region [4.8]
Skin
adius
Superficial
Ulna
Subcutaneous tissue and
Figure: 4.7(b): Anatomical description of human wrist [4.8]
4.6
Concluding Remarks
Presented in this chapter are essential descriptions and dimensional details of humanfinger and wrist. Based on these details, pursued in the next chapter is an effort on
multilayered dielectric modeling of the work-part and relevant analysis of microwave
absorption characteristics is presented.
39
CHAPTER V
MULTILAYERED, LOSSY-DIELECTRIC CYLINDER REPRESENTATION OF
HUMAN FINGER AND THE HUMAN WRIST
5.1
Introduction
Commensurate with the research under study, presented here is an electromagnetic model
of the work-part (human finger or wrist, for example), where the noninvasive sensing
arrangement is intended to be housed. The model indicated thereof is based on the
anatomical formats of the finger and the wrist described earlier in Chapter IV.
A common method of analysis (both for the finger and wrist) is described in the
following sections.
5.2
Evaluation of EM Absorption Characteristics of Human Finger and/or Wrist
Suppose a single layered EM radiating structure is provisioned on the work-part (finger
or wrist) over a length le as shown in Figure 5.1. It is presumed to be excited by an RF
source, say at microwave frequency and another interposed receiving structure senses the
associated interaction. The following assumptions are then valid:
i. The transmitting structure used induces currents in the work-part and in general,
has dimensions comparable or less than the wavelength (A.) of the exciting EM
energy.
40
ii. Any radiation adjoining the transceiver housed on the work-part is suppressed by
a surrounding shield. Hence, in formulating the H-field intensity from the vector
potential induced, it is not necessary to regard the potential; and, any phase delay
due to finite velocity of propagation of EM waves can be neglected [5.1].
iii. The analysis therefore, would be partly quasi-static in nature [5.1].
The analysis in hand refers deducing the formulation that assesses the sensed EM energy
as a function of the dielectric property of the composite multilayered work-part of Figure
5.1. For this purpose, the EM filed components inside the multilayered structure are first
indicated following the results due to Neelakanta [5.1]. The analysis presented in [5.1] is
summarized as follows:
Relevant model is based on equipotential concept. It is assumed that each layer section of
the multilayered structure is equipotential, and therefore, the lateral flow of current is
neglected. This assumption is justifiable due to the fact that, for high conducting tissues
or biomedia at high frequency spectra, the potential gradient exists mainly along axial
direction of the model. Hence the flow of electric current is primarily axial.
41
Also, when the wavelength is large or comparable to the physical size of the work part
plus the shielding, the absence of radiating field component, will sustain no appreciable
gradient in the transverse direction. As a result the impedence of each layer parallels with
those of the other layers [5.2].
The excitation of a multilayered cylinder by an external EM source can be considered as
an extension of a line source exciting a flat multilayered structure. That is, referring to the
Figure 5.2, the biomedia is represented by multilayered dielectric stacking (two
dimensional) and excitation structure refers to a line source and its image placed over the
layers. By the priniciple of reciprocity, any receiving structure improvised can be
modeled likewise.
42
Line source: y-directed current
sheet at z = +h over the
multilayered media occupying z
<0
ffl5€i,At0:/i
X
02, e2, Mo: 4
Image Plane
. A
Viy£s,:Ht>
Image Plane
Image of the line source
Figure 5.2: Two dimensional representation of the work part
43
The electrical characteristics of the biological parts are tabulated in the Table 5.1
Table 5.1 Electrical characteristics of biological parts [5.3]
f
Mhz
1.00
3.00
5.00
10.00
27.12
40.68
50.00
Muscle, blood, skin
ffc S/m
0.400
0.450
0.500
0.625
0.612
0.693
0.722
cc
2000
1600
1200
160
113
97.3
93
Red marrow
ffc S/m
0.22
0.22
0.22
0.22
0.27
0.27
0.27
Cc
100
74
63
40
32
27
25
Bone, Fat, Tendon
ffc S/m
0.03
0.03
0.03
0.03
0.03
0.03
0.03
ec
20
20
20
20
20
14.6
13.4
A comprehensive analysis developed by Neelakanta and presented in [5.1] leads to the
following results.
The EM excitation and the resulting EM sensing can be specified in terms of the
associated power components as follows
Pm= Power dissipated in the lossy work-part
Pt= Power transmitted off from the region due to blood- circulation.
The effectiveness of the EM sensing, therefore is dependent on an efficiency parameter,
(Pm+Pt)
For the line source excitation model presumed, it is indicated in [5.1] that the excitation
structure be sized and placed optimally over the work-part. For example, if a patch
antenna [5.4, 5.5] is chosen, the patch should be kept slightly elevated from the surface of
the work part by an optimal height hoPt= LC(F) - rh , where Lc is the axial length of the
44
antenna; F is a factor dependent on the ratio of Lc and the axial length of the work- part
involved; and rh is the radius of the work-part.
Similarly, for any compatible antenna, its relative disposition (height) from the work-part
(say finger) has to be adjusted for optimum performance.
5.3
Inferential Remarks
The inferences that can be gathered from the above are as follows
•
The geometry of the EM radiating/sensing structures would play a significant role
in deciding the sensing efficiency
•
In dimensioning of such structures the major factors to be taken into account is
the ratio of the size of the structure antenna length to the work-part diameter
•
The sensing structure can be slightly kept elevated from the work-part for
optimum coupling of sensed EM energy.
In view of the aforesaid analytical considerations, the present study is done to evaluate
some typical EM structures compatible for a thimble configuration on the finger (workpart). Relevant details are presented in the next chapter.
5.4
Concluding Note
This chapter is written to offer some theoretical considerations and analytical
formulations for the design of a thimble-housed EM excitation and sensing arrangement
on a finger. The results are duly considered and used in the experimental studies
performed as described in the ensuing chapters.
45
In essence this chapter describes a model that depicts the possible response of the workpart to EM excitation. Relevant response characteristics are useful for the purpose of
evaluating the blood-glucose influx in the work-part.
46
CHAPTER VI
EXPERIMENTAL STUDY AND RESULTS
6.1
Introduction
The experimental efforts carried out as part of this research refers to making some
preliminary investigations as a proof-of-concept of the thesis objectives. The major scope
of this work as reflected in the title of the research is concerned with the eventual
development of a totally noninvasive approach to assay the glucose content in human
beings in vivo.
In the context of the aforesaid scope, addressed is certain prelude of topics indicated in
the previous chapters. In a nut-shell they refer to: (i) Knowing the blood characteristics in
terms of its dielectric properties; and, (ii) identifying compatible work-parts in the human
anatomy wherein the blood characteristics (specifically the glucose content) can be
sensed noninvasively via dielectric characterization.
The proposed effort thereof, involves assessing the blood glucose level by a noninvasive
sensing arrangement. Hence as indicated in the previous chapters, by applying radio
frequency (RF) energy (such as ISM band) external to the work-part, the resulting
electromagnetic absorption can be observed; this EM power absorption is correlated to
the glucose content.
Accordingly, the experimental study undertaken refers to following efforts:
47
•
A proof-of-concept study in the laboratory involving no biological substances or
human subjects. That is, the blood is emulated biomimetically with non-biological
substances; and relevant high-frequency instrumentation is used to assess the EM
absorption properties of the test-material containing varying levels of glucose.
Also, a compatible high frequency transceiver system is conceived toward
realizing the eventual non-invasive sensor arrangement.
•
Further on the basis of the present study, the following efforts are indicated for
further investigations (not studied in the present research):
o The second level of futuristic effort should involve using cell cultures in
vitro depicting the human blood and this can be performed in FAU in
collaboration with College of Science [6.1].
o The third phase should include a prototype structure for clinical
evaluation.
o Lastly, the prototype has to be tested in clinical trials.
6.2
Experimental Efforts of the Present Study
Inasmuch as the present study only refers to the proof-of-concept level, the relevant
experimental efforts carried out are confined to the tasks enumerated and explained
below:
•
Identification and preparation of non-biological substances to make the recipe for
a composite material that represents the biophantom of human blood. That is, the
composite material is designed to portray approximately the dielectric permittivity
and lossy characteristic of human blood
•
Making of a bio mimetic structure representing the human finger using the nonbiological composite synthesized. This effort leads to multilayered unit
representing the bio-phantom of the human finger
48
•
Microwave excitation of the phantom to assess the EM power absorption as a
function of measured glucose added to the phantom. This procedure involves an
EM transmission plumbing at the frequency used (namely X-band or 10 GHz)
•
Design of a transmit-receive antenna system for use on the biomimetic finger
structure and experimental evaluation.
The associated measurements as above involve a set of necessary instrumentation and
equipment as described in the sections that follow.
6.3
Biomimetic Blood: Design and Fabrication
This refers to making of a composite material depicting the phantom of human blood.
The material so designed should conform to the relative permittivity and loss-tangent of
the actual blood. In concocting this bio-phantom material, relevant target values of the
blood's constituents are chosen as per international standards as indicated in [6.2]. The
material synthesized conforms to the standards in the upper frequency band of the EM
spectrum. The permittivity and conductivity of the parameters are indicated in Table 5.1
Table 6.1: Values of permittivity and conductivity at the respective frequencies
according to the Standards of EN 50361 and IEEE 1528-200x
Frequency
(MHz)
E N 50361
IEEE1528-200x
a S/m
€'
a S/m
€'
900
42.3
0.99
41.5
0.97
1800
2100
40.1
1.38
40
1.4
39.6
1.57
39.8
1.49
2450
39.3
1.84
3000
39
2.4
39.2
38.5
1.8
2.4
49
The phantom as suggested and synthesized in [6.2] consists of mineral oil and water with
Triton X-100 (polyethylene glycol) as surfactant. The suitable proportion used in the
recipe is as follows:
•
De-ionized water:
61.3%
•
Mineral oil:
12.6%
•
Triton-X 100:
25.4%
•
Sodium Chloride:
0.70%
Following the steps given in [6.2], the blood phantom is fabricated by heating the mixture
of water-Triton-XlOO-oil very slowly (reaching 45 °C in about fifteen minutes). It is then
homogenized in two minutes.
6.4
Biomimetic Structure of Human Finger: Design and Fabrication
Consistent with the approximate dimensions and anatomical features of an adult human
finger described in Chapter IV, the corresponding biomimetic phantom of human finger
is constructed as illustrated in Figure 6.1.
50
Glass rod of
diameter 5mm
(Emulates the
phalanges)
A
Glass test-tube of
thickness 2 mm
Emulates the
skin/fat
Length : 100 mm
Blood-tissue
composite
phantom
->
Diameter (outer): 13 mm
Figure 6.1(a): Biomimetic model of human finger constructed
Shown in Figure 6.1(b) is the photograph of the finger fabricated. The test-tube
containing the phantom is housed in a plastic syringe for convenient handling during
measurements (the plastic syringe is almost a loss-free dielectric and will not influence
EM absorption by the phantom).
51
fr*
Figure 6.1(b): A photograph of the constructed biomimetic model of human finger
6.5
Microwave Transmission-line Plumbing to Measure Glucose-based Power
Absorption in the Biomimetic Phantom
The proposed work, as stated in the earlier chapters, refers to the development of noninvasive blood-glucose system at the ISM band of 2450 MHz. However, currently due to
non-availability of 2450 MHz source and plumbing, an X-band (10 GHz) system is
adopted instead. Since the study is based on the assumption that the glucose in blood can
approximately be correlated to EM power absorption in the microwave traversing the
52
sample, use of X-band in lieu of 2450 MHz can still prove the concept of the effort
without any loss of generality (except for some relative change in EM absorption).
Consistent with the X-band being adopted, the relevant plumbing and measurement
arrangement is shown in Figure 6.2(a)
-w-
Differential
Amplifier
Detector
1 kHz Square
wave
Modulator
Model:
HP3312A
Function
-WX band source
Model:
Central Scientific
Company 36811
3cm microwave
apparatus
B
Detector
Output Display
Model: Beckman
Industrial Digital
Multimeter
Figure 6.2(a): Microwave (X-band) transmission plumbing: (A) Reference biomimetic
finger with zero glucose content. (B) Test Biomimetic finger with provision to add
glucose incrementally
The photograph of microwave plumbing of Figure 6.2(a) is presented in Figure 6.2(b).
53
-.i?i , W-***'-''-.a
MH
<H
%
ttFigure 6.2 (b): Photograph of the microwave transmission plumbing arrangement.
Shown in the Figure 6.2(c) is a photograph giving the complete view of the experimental
set up consisting of the source, modulator, microwave plumbing with the biomimetic
finger, differential amplifier circuit and voltmeter to measure the detected power.
Figure 6.2 (c): Experimental set up
54
6.5.1
Test Procedure
The schematic diagram of the measurement setup as shown in Figure 6.2 indicates the
microwave source being split into tow identical arms (using a T arm). In each section of
the arm, a microwave-compatible housing unit is included as a part of the waveguide
plumbing. This housing-unit has the provision to hold an inserted bio mimetic finger at
the region of field maximum across the waveguide cross-section. Shown in Figure 6.3 is
the relevant housing unit.
Figure 6.3: Housing unit used to place the biomimetic finger in the experiment.
The test procedure involves the following steps:
•
First, the detected power from each arm (without the biomimetic fingers) is
observed differentially as shown in Figure 6.2(a)
•
Then, in each arm, identical biomimetic fingers are inserted with zero glucose
content
•
Any differential common-mode output is leveled to zero or to a convenient
reference level
55
•
Retaining one test-finger in one of the arms of the microwave plumbing as a
reference unit, aqueous solution prepared with different pre-weighed amounts of
glucose is added to the second arm containing an identical phantom finger.
Addition of glucose is done in incremental steps of 10 /xl of glucose up to 360
mg//il (The corresponding cumulative concentration of glucose is noted down).
Hence the differential output across the two waveguide arms is noted. The
readings are normalized with respect to the reading at zero glucose content
measurement. The concentration of glucose versus EM absorption is tabulated in
Table 6.2. The corresponding plots are presented in Figure 6.4.
56
0
20
30
40
50
60
70
80
90
100
110
121.818
135
147.692
160
173.333
0
20
40
60
80
100
120
140
160
180
200
240
280
300
320
360
Actual
concentration
of glucose
(mg/dl)
Cumulative
concentration
of glucose
(mg/dl)
Trial Trial 2
(V)
1(V)
0.002
0.003
0.074
0.087
0.183
0.135
0.234
0.244
0.234
0.272
0.299
0.275
0.298
0.269
0.312
0.298
0.324
0.307
0.289
0.353
0.232
0.369
0.252
0.357
0.238
0.365
0.42
0.258
0.246
0.41
0.428
0.257
Mea sured
differ ential
rea<lings
0.0065
0.2410
0.5960
0.7947
0.8859
0.9739
0.9706
0.9706
1
0.9413
0.7557
0.8208
0.7752
0.8403
0.8013
0.8371
Trial 1
0.0070
0.2032
0.3154
0.5467
0.5467
0.6425
0.6285
0.7289
0.7570
0.8247
0.86215
0.8341
0.8528
0.9813
0.9579
1
Trial 2
Normalized values
-47.959
-16.595
-8.7304
-6.2316
-5.288
-4.466
-4.4951
-4.4951
-4.2366
-4.7614
-6.6696
-5.9514
-6.4479
-5.747
-6.1607
-5.7807
Trial 1
-44.437
-15.189
-11.373
-6.5951
-6.5951
-5.1928
-5.3844
-4.0963
-3.7685
-3.0239
-2.6389
-2.926
-2.7335
-1.5144
-1.7237
-1.3505
Trial 2
Expressed in dB
Table 6.2: Results obtained using the experimental set up of 6.2 (c)
1
0.30
0.25
0.20
!1
0.15
A
v
Ti-inl 1 !
i run i i
0.10
0.05
0
50
100
150
200
150
200
Concentration of glucose
in mg/dl
50
100
Concentration of glucose
in mg/dl
Figure 6.4(a): Graphs showing the variation of microwave absorption with changing
glucose concentration in two sets of trials performed.
58
c
u
3
04
60
80
100 120
Concentration of
glucose in mg/dl
140
160
*•
180
Figure 6.4(b): Normalized microwave absorption as observed in the two experimental
trials
7
10
Glucose concentration
in mg/dl
12
15
Figure 6.4(c): Plots showing the observed normalized microwave absorption in terms of
relative microwave power in dB
59
121.818
135
240
360
320
280
300
160
173.333
147.692
110
200
160
180
80
90
100
38.8
37.7
35.2
32.2
28.3
32.8
22.8
26.6
28.2
19.8
70
140
12
18.9
40
50
60
6
30
40
60
80
100
120
-10
-0.6
3.5
0
20
0
20
Trial 3
(mV)
Actual
concentration
of glucose
Cumulative
concentration
of glucose
initial reading
32
34.1
41.8
38.4
0.996
0.974
0.924
0.864
0.786
0.876
37.1
39.2
0.676
0.752
0.784
0.616
0.46
0.598
0.34
0.016
0.208
0.29
Trial 3
30.6
33.6
36.5
27.3
19.6
26.5
17.1
12.3
0.8
2.2
Trial 4
(mV)
0.71111
0.85333
0.75777
0.92888
0.87111
0.82444
0.68
0.74666
0.81111
0.60666
0.38
0.43555
0.58888
0.27333
0.04888
0.01777
Trial 4
Normalized values
-0.034813
-0.228821
-0.686561
-1.269725
-2.091549
-1.149918
-3.401066
-2.475643
-2.113679
-4.208386
-6.744843
-4.465976
-9.370422
-35.9176
-13.63873
-10.75204
Trial 3
-2.4091626
-2.9612507
-0.640724
-1.3776257
-1.1985289
-1.676772
-2.5374647
-1.8183929
-4.3409973
-3.3498217
-8.404328
-7.2191288
-4.5993327
-11.266148
-26.215796
-35.00245
Trial 4
Microwave absorption
in dB
Table 6.3: Set of experimental results obtained using a single biomimetic finger: Measured differential results with respect to
0
20
40
60
80 100 120 140
Glucose concentration
160
•
180
200
Figure 6.5(a): Plots of microwave absorption with changing glucose concentration in
two trials
60
80 100 120 140
Glucose concentration
in mg/dl
160
180
200
Figure 6.5(b): Plots of normalized microwave absorption with changing glucose
concentration in two trials
61
T
i
i
1
r
i.
o
a
10
1
i
i
- i-
1
i
i
-i
>
i
i
f
C3
i
h
1
i
J ?
Jr
^ Q L ^ ^
f
\
1
\
i
i
1
\
-i
1
\
f
i
u
t
i
i
a
i
i
i
i
i
i
1
\
I
I
u
J-
\
i
i
i
i
i
i
i
i
i
r
i
^£JST
i/"^ T
j
V
I
~*~~ Trial
I «
I.S-20
>
&
PS
-•-Trial
! /
-30
I
i / X
11/
0
/
!
!
!
!
i
i
i
i
i
i
i
i
i
i
i
i
;
i
2
4
[
6
8
10
12
14
16
18
20
Glucose concentration
^
in mg/dl
Figure 6.5(c): Plot of normalized relative microwave absorption in dB with changing
glucose concentration in two trials
The results obtained in the two experiments (as shown in the tables 6.2 and 6.3) are
further used to plot the graph shown in the Figure 6.6. The logarithm of the normalized
readings is used to calculate the average and then, the anti-logarithm of the average is
plotted to obtain the expected trend of the clustered data. The Table 6.4 is presented to
show relevant calculations.
62
Table 6.4: Normalized readings obtained from the Tables 6.2 and 6.3 to plot graph in
Figure 6.6
Actual
concentration
of glucose
0
20
30
40
50
60
70
80
90
100
110
121.818
135
147.692
160
173.333
L(T1)
-5.033
-1.422
-0.517
-0.229
-0.121
-0.026
-0.029
-0.029
0.000
-0.060
-0.280
-0.197
-0.254
-0.173
-0.221
-0.177
L(T2)
-4.960
-1.593
-1.153
-0.603
-0.603
-0.442
-0.464
-0.316
-0.278
-0.192
-0.148
-0.181
-0.159
-0.018
-0.043
0.000
L(T3)
-4.135
-1.570
-1.237
-1.078
-0.776
-0.514
-0.484
-0.391
-0.285
-0.243
-0.132
-0.240
-0.079
-0.146
-0.004
-0.026
L(T4)
-4.029
-3.018
-1.297
-0.967
-0.831
-0.529
-0.499
-0.385
-0.292
-0.209
-0.193
-0.138
-0.073
-0.158
-0.277
-0.340
L(Tx): Logarithm of normalized readings of trial x (=1, 2, 3, 4)
63
Average
-4.539
-1.901
-1.051
-0.720
-0.583
-0.378
-0.369
-0.280
-0.213
-0.176
-0.188
-0.189
-0.141
-0.124
-0.136
-0.136
Antilog
of
average
0.010
0.149
0.349
0.486
0.558
0.685
0.691
0.755
0.807
0.838
0.828
0.827
0.867
0.883
0.872
0.872
Concentration of glucose
in mg/dl
Figure 6.6: Plot of the average values of the readings obtained in the four trials
elucidated via logarithmic clustering (as indicated in Table 6.4)
64
6.6
Noninvasive Microwave Transmit-receive Sensor Placed on the Biomimetic
Finger: Design Details and Measurements
This measurement is concerned with developing a structure that can house the phantom
finger and measure the changes in the glucose content added to the phantom. This
essentially is the proof-of- concept structure that represents the eventual thimble-like
non-invasive sensor needed in practical applications. The test structure in question
essentially consists of a dielectric tube within which the phantom can be inserted. On the
surface of the dielectric holder (of low-loss plastic material) a compatible set of
transmit/receive metallic patch antennas are mounted. Further, in order to emulate
micro strip-like antenna configuration, a layer of high permittivity material (such as TiCh)
is circumscribed over the patches. Lastly, a ground-plane around the surface is facilitated
to complete the required configuration. The overall geometry of the structure is presented
in the Figure 6.7 and the corresponding photographs are presented in the Figure 6.8
65
Front elevation
Side elevation
a
b
c
d
•
T
A
A
receiver
From
transmitter ^
*
Section on 'AA'
Figure 6.7 (a): Patch antenna assembly, a - Low loss plastic tube housing the phantom
circumscribed around the patch antenna; b, c- parasitic elements of the antenna structure;
d- transmit/receive elements of the patch antenna
The patch antenna set is mounted on a low-loss plastic tube that houses the test-phantom.
The antenna set consist of four units: Two square patches placed face-to-face on
diametrically opposite sides. One of them is connected to the transmitter and the other to
the receiver. The pair of rectangular patches are parasites included to enhance the EM
interaction with the phantom.
66
Cut-sectional view
with enclosure
Figure 6.7(b): A cross sectional view of the antenna structure covered by an enclosure.
GP- Ground Plane; D-dielectric made of high permittivity material such as TiCh; PAPatch antenna mounted on a plastic holder (PT) that houses the phantom assembly
Indicated in the Figure 6.7(b) is an enclosure placed on the patch antenna assembly
shown in Figure 6.7(a). The enclosure consists of circumferential layer of high
permittivity material (TiCh) and a ground plane both overlaid on the patches (mounted on
a plastic holder). In effect, the patch antenna plus the holder unit emulates the conditions
approximately of a microstrip radiator.
67
Figure 6.8: Photograph of the phantom housed in the thimble-like enclosure
6.6.1
Patch Antenna Assembly Details
The patch antenna assembly illustrated in the Figure 6.6 consists of two simple square
patches placed face-to-face on a plastic tube. One patch is connected to the source and
another is connected to the detector/receiver. Interposed between the patch antennas on
the plastic tube are diametrically opposite rectangular patches unconnected to the
transmitter/receiver. These are parasitic elements.
68
The dimensions of the patch antenna used are compatible to the dimensions of the tube
that houses them relative to the wavelength. The criteria in choosing the type of the
patch, relevant dimensions and the disposition of the patches on the cylinder are as
follows:
•
The shape and size of the patches should be such that the transmitting and the
receiving parts should remain well isolated; and, any coupling between them
should be negotiated only through the phantom. This would allow maximum
interaction of the EM energy on the phantom so that, the corresponding power
absorption will depict the lossy nature of the phantom constituent (such as the
glucose content)
•
The parasites included enhance the said interaction in a multiple manner with the
material underneath so as to increase the sensitivity of the observed artifact
•
The high dielectric substrate improves the patch antenna performance [6.3, 6.4]
•
Several versions of the patch antennas were fabricated and tested for acceptable
performance. They are illustrated in the Figures 6.9. The final structure is as
shown in the Figure 6.9(a) and its selection was based on its performance. Yet,
this structure may not be an optimum structure, but it indicates an approach
towards a conceivable design and is used in the present study.
69
Figure 6.9(a): The patch antenna eventually chosen for the experiment based on its
performance: Shown is the patch antenna unit and the same patch antenna with the
grounded-plane shield
Figure 6.9(b): Photographs of the other antennas tested
70
6.6.2
Patch Antenna Details
Using the transmit-receive system of Figure 6.6, the absorbed EM power across the
phantom as a function of the glucose content is measured. Two types of measurements
were taken
1. With varying amounts of glucose in saline solution;
2. With varying amounts of glucose in aqueous solution
The measured differential changes in normalized form are tabulated in the Table 6.5 and
Table 6.6. The relevant plots are presented in the Figures 6.10(a) and 6.10(b).
71
Table 6.5: Measurements using the patch antenna with varying amounts of glucose in
saline solution
Concentration
of glucose
in mg/dl
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
Initial
(V)
0.173
0.173
0.173
0.173
0.173
0.173
0.173
0.173
0.173
0.174
0.174
0.174
0.173
0.175
0.174
0.174
0.175
0.175
0.175
0.175
0.175
Final
(V)
1.846
1.823
1.819
1.865
1.705
1.727
1.819
1.795
1.801
1.787
1.785
1.707
1.75
1.706
1.783
1.782
1.733
1.767
1.729
1.703
1.713
Reading
Difference
Normalized reading
(V)
1.673
1
1.65
0.98625
1.646
0.98386
1.692
1.01136
1.532
0.91572
1.554
0.92887
1.646
0.98386
1.622
0.96952
1.628
0.9731
0.96414
1.613
0.96294
1.611
0.91632
1.533
0.94262
1.577
1.531
0.91512
1.609
0.96175
1.608
0.96115
1.558
0.93126
1.592
0.95158
1.554
0.92887
0.91333
1.528
0.91931
1.538
72
Table 6.6: Measurements using the patch antenna with varying amounts of glucose in
water
Concentration
Reading
of glucose
Initial Final
in mg/dl
(V)
0
(V)
Diff
Normalized
(V)
reading
0.145 0.145
0
0
20
0.145
1.655
1.51
0.8861
40
0.139
1.632
1.493
0.8761
60
0.141
1.725
1.584
0.9295
80
0.14
1.657
1.517
0.8902
100
0.139
1.679
1.54
0.9037
120
0.136
1.84
1.704
1
140
0.135
1.696
1.561
0.9160
160
0.133
1.585
1.452
0.8521
180
0.133
1.598
1.465
0.8597
200
0.132
1.726
1.594
0.9354
240
0.132
1.643
1.511
0.8867
280
0.133
1.586
1.453
0.8526
300
0.133
1.686
1.553
0.9113
320
0.133
1.595
1.462
0.8579
360
0.133
1.763
1.63
0.9565
UB
1.0
Im
x ^
l_
0.9 "^ " ^ L v
1
* * •
-b T -~ " T=-"
'
'"••!•.
DC
5 5 0.9
'*.._
•8 |
'flHMi
T #^
__y^J-'--L-^.
0.9
!
T
Banana
*t*»,^
r--. t>
F"*-.
!_ '"•^••- """•
5
10
Concentration of glucose
in mg/dl
Figure 6.10(a): Microwave absorption versus glucose concentration in saline solution
using patch antenna arrangement (data from Table 6.4); UB: upper bound; LB: Lower
bound; Im: Infimum; Sm: Supremum
1.0
<=»»*^l
/k^/^T
0.8
wo
0.6
//
1 _.
! ..
i
L
i
i
j
Polynomial trend
1
1
1
1
1
1
1
1
1
1
1
1
I
--•—Measured values
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
S
I
Z §0.2
1
o
50
100
1
I
2&
150
1
200
250
1
300
1
350
-i
400
Concentration of glucose
in mg/dl
Figure 6.10(b): Microwave absorption versus glucose concentration (in water) using
patch antenna arrangement (data from Table 6.5)
74
Shown in the Figure 6.11 is the experimental arrangement used.
•
.-jJwNd* *m
IHIMi
Figure 6.11: Experimental set up with patch antenna based thimble arrangement
6.7
Discussions on the Results
6.7.1
General Observations
The experimental studies carried out provide a general confirmation on the feasibility
aspects of the conceived sensor. The results indicate that varying extents of glucose-
75
content in the blood-phantom vary the EM (microwave) absorption. Relevant precursive
data obtained follow the itemized details presented below:
•
Test Set I: Figure 6.4/Table 6.2
The first test results illustrate the data from two separate trials. In trial 1,
there appears to be a discontinuity (around 100 mg/dl) of glucose. It is
considered as due to some asymmetry in the geometrical/structural
dispositions of reference and test units. Therefore a second trial (trial 2)
was attempted and the discontinuity seen is reduced. However, in both
trials, the measured EM absorption monotonically increases with glucosecontent. That is:
•
For low concentrations of the glucose content, the EM absorption
varies linearly; but at larger concentrations (> 100 mg/dl) the
variation shows nonlinearity.
•
The relative EM absorption is about 6 dB for a change from 0 to
150 mg/dl of glucose (see Figure 6.4(c))
•
Test Set II: Table 6.4/Figure 6.10
This corresponds to the measurements done via patch antennas mounted
on the biophantom. The results indicate:
•
Variation of measured EM output with respect to glucose content
•
Feasibility of sensing EM absorption noninvasively via patch
antenna arrangement compatible for eventual design of a thimblelike structure.
6.7.2
Theoretical Results
In Chapter III, Section 3.4, the dielectric model of blood with varying levels of glucose is
discussed. It is indicated that a fractional change in glucose content will be seen as a
fractional change in the loss angle of the EM absorption.
76
Such fractional change relation implicitly specifies a nonlinear dependence of the
variables. The experimental results confirm this nonlinearity.
Further in Chapter V, it is shown with a multilayered model of finger that the biological
dielectric characterization within the finger (due to say, glucose content) can be sensed
external to the model via suitable EM transmitter-receiver system. The present
experimental studies indicate positive directions thereof.
6.8
Inferential Remarks
From the theoretical considerations and experimental studies performed , the following
inferences can be made:
•
Blood-glucose content alters the dielectric characteristics of the blood
•
Corresponding dielectric loss can be assessed via suitable EM absorption
measurement
•
Compatible frequencies suggested here are 2450 MHz and 5 GHz of ISM band
•
A patch antenna assembly can be devised to form the thimble arrangement for
noninvasive assessment of blood-glucose content
•
However, such antenna assembly has to be optimized with more extensive studies
•
EM absorption by blood may also depend on other constituents of blood (such as
bile, urea, RBC, etc) as observed by Neelakanta et al. [6.5].
Therefore, the
present study has to be more comprehensively addressed to segregate the artifacts
due to various blood constituents
More on the related issues and scope for further studies are indicated in the next chapter.
6.9
Concluding Note
This chapter is mainly a compendium of details on the experiments performed to
ascertain the scope of the thesis. This is only a preliminary effort. Nevertheless the results
77
show promising directions. Also, the experimental data concur with some theoretical
observations given in the earlier chapters. Further, the tests performed indicate the
approach toward realizing a compatible transmitter-receiver unit for noninvasive bloodglucose assessment.
78
CHAPTER VII
WHAT LIES AHEAD?
7.1
Introduction
The present study is largely an experimental attempt to verify the conceived hypothesis
on sensing the glucose content in blood (stream) noninvasively across a human
anatomical part. The underlying concept relies on the fact that an EM wave rendered on
the work-part will be attenuated because of the absorption suffered due to lossy dielectric
nature of the material. It is proposed in this study that the lossy nature can be correlated
to the glucose-content in the material; hence the evaluation of EM absorption can be
linked to the glucose level within the material.
Thus, the research addressed is essentially a proof-of-concept experimental effort with
some underlying theoretical heuristics. This chapter gives a critical outline on the
promises of feasibility to pursue this research and indicates open-questions that lie ahead
in any such pursuits.
7.2
Underlying Hypotheses of the Present Study
As indicated above, the essence of the present work resides on its gist of hypothesis,
namely:
•
Glucose content alters the dielectric properties of blood
79
•
Such altered dielectric artifact can be sensed via EM absorption by the blood
being tested
•
Relevant EM absorption can be non-invasively assessed by a suitable
RF/microwave arrangement.
Now, one can critically appraise the hypotheses as above in view of the available details
in the literature and based on the data collected in the present study.
7.2.1
Glucose Content Alters Dielectric Property...
This statement has been proven in the existing studies [7.1] as well as from the
observations of the present work:
•
In [7.1], it is clearly demonstrated that dielectric properties of blood-imitators
change with glucose content. Relevant study is performed in the millimenterwave range EM spectrum
•
The work by Liao et al. [7.2] supplements the above inference based on the results
obtained at 2450 MHz
•
The theoretical model by Neelakanta et al. [7.3] again shows how blood being a
statistical mixture of several constituents (such as glucose) will have a
corresponding dielectric characteristics as decided by the volume fractions of the
constituents
•
Results of study as presented in the Figures 6.4, 6.5 and 6.9 also confirm the
change of the dielectric property of a blood-phantom as a function of glucose
content.
7.2.2
Critical Questions
While it can be accepted that a change in the concentration of a constituent can alter the
dielectric features of blood, the lingering question is how to separate the distinct
80
contributions to the dielectric property from each of the constituents. For example, it has
been shown by Neelakanta et al. [7.4-7.6] that hematocrit value (RBC content) of blood
can also affect the conductivity (and hence the dielectric properties) of blood. If so, by
observing EM absorption (due to dielectric loss), one has to segregate the relevant effects
due to say, glucose and hematocrit value. Likewise other constituents (such as bile
content, urea etc. in blood) may also affect the EM absorption profile [7.4-7.6]. Relevant
issues need careful consideration.
7.3
Any Percentage Change in Blood Constituents can be Sensed via EM
Absorption Measurements
The above statement is a valid proposition as verified in the works of [7.1]. The present
study also supplements this observation.
7.3.1
Critical Questions
In the efforts of [7.1], the experimental study refers to an invasive effort of collecting
blood sample and subjecting it to a laboratory study in order to assess the constituent
percentage. Such methods are not therefore, suitable for a non-invasive approach.
The present study on the other hand indicates a non-invasive strategy with promising
results. However, the sensitivity of noninvasive interception should be improved with
careful design of the transmitter-receiver system.
81
7.4
Directions for Future Study
Consistent with the details of Section 7.2 and 7.3, the present work indicates a
preliminary scope with a basic proof-of-concept, which can be extended after
incorporating the following:
•
The non-invasive structure should be optimized for reliable and robust operation
and for improved sensitivity. Necessary redesigns should be done and
investigated
•
Segregation of dielectric absorption artifacts due to different contents of the blood
should be carried out. As mentioned before, not only glucose, but also RBC count
could affect the EM absorption and hence would alter the dielectric property of
blood [7.4-7.6].
Further, in a recent study, it has been indicated that the presence of HIV/AIDS virus in
the blood may also indicate characteristic EM absorption profiles [7.7].
Therefore, it is suggested here that any instrumentation (invasive or non-invasive) that is
used to assess the blood glucose content must be able to segregate the EM absorption
profiles due to different blood constituents. Such revaluation/redesign options may refer
to:
•
Using simultaneously two distinct (allowed) frequencies in the test procedure: For
example, ISM bands of 2.45 GHz and 5 GHz can be adopted in the
instrumentation. If such non-invasive method is pursued, suitable patch antennas
(dual band) can be designed at ease [7.8, 7.9]
•
Artificial intelligence can be adopted for the required segregation and decision
making. This intelligence can be a part of associated chip design
•
Dielectric absorption spectra of blood constituents should be thoroughly
investigated. This will help not only the glucose abnormality but also would
82
facilitate screening for other pathological conditions such as HIV/ AIDS, sickle
cell anemia etc.
•
The observed dielectric absorption versus glucose concentration in the test sample
is nonlinear at larger glucose levels. Therefore, necessary linearization may be
required in the electronic circuits before display.
•
Similar to finger-prick glucose test procedure, periodic calibration with a standard
solution is recommended. Relevant calibration considerations should be studied.
•
Another promising future effort could be using EM technique on the invasive
finger-prick method. That is, the test strip mounted with a capillary can access the
finger-prick blood and it can be subjected to EM absorption test for sensing the
glucose content. Since the strip has no proprietary chemicals contained in it, it can
be made at a relatively low cost. Here again, multiple frequency strategy can be
adopted to distinguish glucose from other contents.
7.5
Closure
This research should be regarded as a preliminary step in assessing the blood
characteristics via noninvasive approach. Nevertheless, it indicates a scope and offers
future direction with open questions. It is expected to culminate into a fruitful
contribution to the society.
83
CHAPTER VIII
EXECUTIVE SUMMARY
The research addressed in this thesis is an effort motivated to seek a method of assessing
the glucose-content in the human blood noninvasively. That is, unlike the traditional
invasive (finger-prick) method of testing for glucose level in blood, attempted here is a
proof-of-concept to develop a technique that can non-invasively probe for the blood
chemistry and estimate the glucose content thereof.
The conceived research in the pursued effort relies on the fact that blood is a
heterogeneous mixture of ionic and molecular contents; and, a change in the volumefraction of a constituent would influence the dielectric property of the whole blood.
Hence, an electromagnetic (EM) device can be conceived to sense the dielectric property
so as to assess the concentration of the corresponding constituents.
The proposed method is based on exciting the whole blood streaming through a human
part (such as finger, wrist or ankle) with EM energy and then sensing the dielectric
absorption characteristically posed by a constituent of interest (proportional to the
volume fraction of the constituent).
Consistent with allowed EM energy band for medical applications, the ISM band is
suggested for this bio-sensing.
The study performed thereof is summarized as follows:
84
•
Biomimetic model of a finger is designed with non-biological substances
•
The blood-imitation is emulated with a mixture of mineral oil, water,
polyethylene glycol and sodium chloride. This biophantom is used to constitute
the test biometric finger
•
The test biomimetic finger is subjected to microwave excitation (at X-band: 10
GHz) and the corresponding EM power loss is sensed as a function of varying
concentration of glucose (from 0 mg/dl to 200 mg/dl) added to the biophantom of
blood. Hence the glucose content is correlated to the EM power absorption
•
To emulate the non-invasive method of sensing the glucose component in blood,
the biomimetic finger is housed in a thimble-like structure with a pair of patch
antennas - one for EM excitation and the other for EM detection (after the EM
energy synergism with the biophantom). Again, EM absorption is measured as a
function of glucose content in the biophantom.
The results obtained from the preliminary experimental studies as above lead to the
inference of viable feasibility of the proposed biosensing. However, it is concluded that
more efforts on this proof-of-concept is needed towards the next step. A prototype can be
designed after this proof-of-concept study is extended to an emulated blood synthesized
via cell cultures.
Moreover, the artifacts of EM absorption due to various contents (other than glucose) in
blood have to be segregated. It is suggested that one of the design solutions to this might
be to use a dual frequency approach (using two ISM band frequencies) in combination
with artificial intelligence.
As a closure, it is indicated that this research is only a precursor and a proof-of-concept
effort. Relevant theoretical considerations performed support the feasibility of the
approach. Further, some existing studies on glucose-based EM absorption support the
85
hypothesis pursued. However, to the best of the author's knowledge, no noninvasive
biosensing as conceived in this study has been reported in open literature or patent
documented. It is surmised that this research could culminate in a useful device for
management of diabetes.
86
REFERENCES
CHAPTER I
[1.1]
Homepage of Zigbee™ alliance: www.zigbee.org
[1.2]
Webpage
of
US
Food
and
Drug
Administration/diabetes
http://www.fda.gov/diabetes/glucose.html and webpage of National Diabetes
Information Clearinghouse http://diabetes.niddk.nih.gov/dm/pubs/glucosemonitor
/index, htm
[1.3]
J. L. Smith, The pursuit of noninvasive glucose: "Hunting the deceitful turkey",
http://www.mendosa.com/noninvasiveglucose.pdf, obtained on June 17, 2008
[1.4]
A. Caduff et al., Noninvasive glucose monitoring in patients with diabetes: A
novel system based on impedance spectroscopy, Biosensors and Bioelectronics,
22(2006), 598-604
[1.5]
P.S. Neelakanta, et al., A dielectric model of human blood, Biomed Technik
28(1983), 1 8 - 2 2
CHAPTER II
[2.1]
P.S. Neelakanta et al., A dielectric model of human blood, Biomed Technik
28(1983), 1 8 - 2 2
[2.2]
F. B. Armstrong, Biochemistry, Oxford University Press, New York, NY: 1983
87
[2.3]
Image source: Webpage http://en.wikipedia.org/wiki/Glucose obtained on Feb 12,
2008
[2.4]
Image source: Webpage http://www.fda.gov/fdac/features/2002/102_diab.html
obtained on June 17, 2008
[2.5]
Blood sugar regulation, Wikipedia webpage :
http://en.wikipedia.org/wiki/Blood_sugar_regulation, obtained on June 27, 2008
[2.6]
Christopher D. Saudek, Richard R. Rubin, and Cynthia S. Shump. The Johns
Hopkins Guide to Diabetes. Baltimore: The Johns Hopkins University Press, 1997
[2.7]
Image source: webpage: http://www.clarian.org/ADAM/doc/graphics/images/en
/l 9914.jpg, obtained on April 10th 2008
[2.8]
L.S. Kuhn, "Biosensors: blockbuster or bomb?" The Electrochemical Society
Interface, 7 (4) (1998), pp. 26-31
[2.9]
L.C. Clark, Monitor and control of blood and tissue oxygen tension. Transactions
of American Society for Artificial Internal Organs 2, (1956) 41-48
CHAPTER III
[3.1]
P.S. Neelakanta et al., A dielectric model of human blood, Biomed Technih
28(1983), 1 8 - 2 2
[3.2]
P.S Neelakanta, Handbook of electromagnetic materials: monolithic and
composite versions and their applications, CRC press, 1995, Boca Raton,
FL:1995
[3.3]
K. Lichtenecker
and K. Rother,
"Die Herleitung des
logarithmischen
Mischungsgesetzes aus allgemeinen Prinzipien der stationaren Stromung," Phys.
Zeitschr., 32 (1931) 255-260
88
CHAPTER IV
[4.1]
Image source: http://www.handuniversity.com/topics.asp7Topic_ID=13, obtained
on April 14,2008
[4.2]
Image source: http://en.wikipedia.org/wiki/Hand, obtained on April 14, 2008
[4.3]
Image source: http://en.wikipedia.org/wiki/Deep_palmar arch, obtained on April
15th 2008
[4.4]
Image source: Detailed structure of a finger, Anatomy, Quick Study Reference
Guide, BarCharts Inc., Boca Raton, FL
[4.5]
V. Bundhoo, E.J Park, E.J., "Design of an artificial muscle actuated finger
towards biomimetic prosthetic hands," Proceeds of 12th International Conference
on Advanced Robotics, 2005. ICAR '05, July 18-20 (2005), Seattle, Washington
pp 368-375,
[4.6]
C. D Clemente: Anatomy: A regional atlas of the human body, Urban and
Schwarzenberg, Baltimore MD: 1987
[4.7]
P.S. Neelakanta: Rewarming of hypothermic human-limbs in-vivo by RF
induction heating: Theoretical considerations. Report # II, IGPA Assignment, June
1988- August 198, NAMRL, Pensacola, Florida
[4.8]
H.S.Ho: Energy Absorption patterns in circular triple layered tissue cylinders
exposed to plane wave sources. Health Physics 31 (1976), 97-108
CHAPTER V
[5.1]
P.S. Neelakanta: Rewarming of hypothermic human-limbs in-vivo by RF
induction heating: Theoretical considerations. Report # II, IGPA Assignment, June
1988- August 1988, NAMRL, Pensacola, Florida
89
[5.2]
M. Hoque and O.P. Gandhi: Temperature distributions in the human leg for VLFVHF exposure at the ANSI -recommended safety levels, IEEE Trans. Biomed.
Eng., BME-35(6), (1988), 442-449
[5.3]
H.S.Ho: Energy Absorption patterns in circular triple layered tissue cylinders
exposed to plane wave sources. Health Physics 31 (1976), 97-108
[5.4]
P.S Neelakanta and R. Chatterjee, Antennas for information super skyways: An
exposition on outdoor and indoor wireless antennas, Research Studies Press Ltd.,
UK:2003
[5.5]
K.C. Gupta, et al., Microstrip lines and slotlines, Artech House, Boston MA:
1996
CHAPTER VI
[6.1]
Personal Communication with Dr. R. Narayanan, College of Science, Florida
Atlantic University, May 2008
[6.2]
V. Vigneras and F. Bonnauddin, Biological tissues equivalent liquids in the
frequency range 900 - 3000 MHz, PIOM Laboratory ENSCP, University of
Bordeaux, France
[6.3]
P.S Neelakanta and R. Chatterjee, Antennas for information super skyways: An
exposition on outdoor and indoor wireless antennas, Research Studies Press Ltd.,
UK: 2003
[6.4]
K.C. Gupta, et al. Microstrip lines and slotlines, Artech House, Boston MA; 1996
[6.5]
P.S. Neelakantaswamy, et al., Conductimetric method of determining the
hematocrit value of the blood, Rev. Set Instrum. (USA), 54 (1983), 1186-1190
90
CHAPTER VII
[7.1]
V.V. Meriakri, E.E. Chigrai, LP Nikitin, M.P. Parkhomenko, "Dielectric
properties of glucose solutions in the millimeter-wave range and control of
glucose content in blood," The Joint 30th International Conference on Infrared
and Millimeter
Waves and 13th International Conference on Terahertz
Electronics, 2005. IRMMW-THz 2005, Williamsburg, Virginia USA. vol 1 (2005)
pp. 315-316
[7.2]
X. Liao, et al, Dielectric properties of alpha D-glucose aqueous solutions at 2450
MHz, Food Research International 36, (2003), 485-490.
[7.3]
P.S. Neelakanta, et al., A dielectric model of human blood, Biomed Technik
28(1983), 1 8 - 2 2
[7.4]
P.S. Neelakantaswamy, et al., Conductimetric method of determining the
hematocrit value of the blood, Rev. Sci. Instrum. (USA), 54, (1983), 1186-1190
[7.5]
P.S. Neelakantaswamy, K.F. Aspar, A. Rajaratnam and N.P. Das: A Simple
Conductivity Meter for Applications in Clinical Chemistry, Instrumentation
Science and Technology (USA), Vol. 13(1), 1984, 69-78
[7.6]
P.S. Neelakantaswamy, et al., A conductimetric experiment to assay hemoglobin
in blood, Med. and Biol. Eng. and Comput. (UK), 22, (1984), 367-370
[7.7]
A. Lonappan et al., A novel method of detecting HIV/AIDS using microwaves,
Microwave and Optical Technology Letters, 50 (2008) pp 557-561
[7.8]
P.S Neelakanta and R. Chatterjee, Antennas for information super skyways: An
exposition on outdoor and indoor wireless antennas, Research Studies Press Ltd.,
UK: 2003
91
[7.9]
K.C. Gupta, et al., Microstrip lines and slotlines, Artech House, Boston MA:
1996
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 375 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа