.. £24. .
I ”r’vv Z . .

'3. .1.
if; 753‘

 

 

 

 

I... A. V
‘ 1.
a. .I\.u.r .
. L. 3|

.3: .. .. ,. .‘
14.1. a;

. ., ofv...“ ”.st

. ...-..V.\. (.1
:. , 1,!

 

..1.4§.h.

 

TH???
a
1563

This is to certify that the

thesis entitled

Fluorescent Thermography Methods for
Bio-Thermal Therapies

presented by

Damien J. Fron

has been accepted towards fulfillment
of the requirements for

MS Jegree in Mechanical Engineering

9« WM

Major professor

Date 87/4/01—

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

.UBRARY
Michigan State
University

 

 

 

PIACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuepGSp 15

 

FLUORESCENT THERMOGRAPHY METHODS FOR BIO-THERMAL
THERAPIES

By

Damien J. Fron

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTERS OF SCIENCE
Department of Mechanical Engineering

2002

ABSTRACT

FLUORESCENT THERMOGRAPHY METHODS FOR BIO-THERMAL
THERAPIES

By

Damien J. Fron

Thermotherapy by way of laser and radiofrequency (RF) heating devices has become
ever more popular for the treatment of joint instabilities associated with the knee,
shoulder, wrist, and hip. Similarly, cooling/freezing therapies are of biomedical interest
at the opposite temperature extreme for such applications as cryosurgical ablation of
tumors. This thesis addresses the need for improved methods of temperature
measurement required to understand the performance of various therrnotherapy devices
as well as the thermal response they produce within biomaterials. Two fluorescent
thermography methods of biomedical interest have been developed to measure
temperatures from —80°C to +90°C using temperature-dependent fluorescent
techniques. Temperatures as high as 90°C were recorded within aqueous solutions and
tissue phantoms using two-color laser-induced fluorescence thermography (LIFT) and
the ratio of the fluorescent emission intensities of the dissolved fluorescent dyes
Rhodamine B and Rhodamine 110. The feasibility of using fluorescent chemicals to
measure temperatures for cryogenic application was evaluated using fluorescent
emission ratiometric imaging thermography (FERIT) and the fluorescent emission
intensity of Terbium. Present capabilities of the fluorescent thermography methods
mentioned enable 10 Hz framing rates at an accuracy of 15°C, with a spatial resolution

of 90 um and a temperature resolution of 0.02°C.

ACKNOWLEDGEMENTS

I would like to take a few moments to thank those who have supported me
throughout my graduate career at Michigan State University. First, I would like to
thank my advisor Dr. John J. McGrath. He welcomed me into his lab as an
undergraduate assistant for his graduate students, and because of those experiences and
his encouragement, I eventually decided to pursue my masters degree with him. I
believe that many of my experiences in his labs have turned me into the student I am
today. Dr. McGrath exhibits excellent commitment to his students and has been a
positive role model in my life. I thank him for all that he has taught me these past few
years.

I would also like to thank my guidance committee members Dr. Neil Wright and
Dr. Robert McMasters for their comments and advice on my thesis. The time you spent
with me helped me to explore the fill] scope of my project.

In the three years I worked in Dr. McGrath's lab I had the pleasure of knowing
Dr. Gloria Elliott, Dr. Paul Hoke, and Dr. Alptekin Aksan. One by one they graduated
leaving me by myself under the fiill attention of Dr. McGrath. I thank them for their
friendship and for sharing their knowledge and experiences with me. A special thanks
to Dr. Aksan who taught me: "A day in the lab without disrespect is not a day in the
lab."

In addition to those already mentioned I appreciate the many hours of advice
and discussion from Nate Verhanovitz and Chee Lum. Thank you for your help since it

surely contributed to the completion of this thesis.

iii

I am also thankful to my mom, dad, and brother for their enduring support
throughout the years. I know they thought I would never be done and I could not have
finished without their love. Just think, I could have continued for my Ph.D. Last and
surely not least I thank my love, Sara, who saw me through the ups and downs of my

graduate career and offered endless encouragement and compassion.

iv

TABLE OF CONTENTS

 

 

 

 

 

LIST OF TABLES vii
LIST OF FIGURES viii
KEY TO SYMBOLS AND ABBREVIATIONS xi
Chapter 1: Introduction 1
1.1 Introduction to Radiofrequency Heating .......................................................... 3
1.2 Evaluation of Common Temperature Measurement Techniques ..................... 5
1 .2.1 Thermocouples .......................................................................................... 5
1.2.2 F luoroptic Thermometry ........................................................................... 7
1.2.3 Infiared Thermography ............................................................................. 8
1.3 Introduction to Temperature Measurement at Cryogenic Levels ................... 11
1.4 Fundamentals of Luminescence ..................................................................... 12
1.5 Theory of Temperature Dependent Fluorescence .......................................... 16
1.6 Introduction to LIFT ....................................................................................... 18
1.7 Introduction to F ERIT .................................................................................... 20
Chapter 2: Evaluation of Temperature Dependent Fluorescent Chemicals ...... 22
2.1 Evaluation of F luorophores with a Spectrometer ........................................... 24
2.2 Evaluation of the F ERIT Fluorescent Chemicals ........................................... 28
2.3 Spectral Evaluation of the Polymer Based F luorophores with the Ocean
Optics Spectrometer ....................................................................................... 29
2.4 FERIT Ratiometric Methods for the Determination of Temperature ............. 33
2.5 Temperature Dependent Fluorescence Evaluation of Europium ................... 36
2.6 Temperature Dependent Fluorescence of Terbium ........................................ 39
2.7 Evaluation of the LIFT fluorescence chemicals ............................................. 46
2.8 Evaluation of the Spectral Characteristics of Rhodamine B and 110 ............ 47
2.9 LIFT Ratiometric Methods for the Determination of Temperature ............... 49
2.10 Temperature Dependence of Rhodamine B, Rhodamine 110 Solution ......... 51
2.11 Fabrication of 488nm Light Source for Calibration of Rhodamine B-110
Solution .......................................................................................................... 54
2.12 Effect of Concentration of Rhodamine B-110 Solution on Spectral Emission
........................................................................................................................ 56
Chapter 3: Application of LIFT and FERIT to Measure Temperature Fields On
and Around Radiofrequency Heating Devices 59
3.1 Application of LIFT to Measure Temperature Fields Surrounding RF Probes
........................................................................................................................ 59
3.2 Image Acquisition Rate Improvement ........................................................... 63
3.3 Field Temperature Measurements Around RF Probes ................................... 66
3.4 Results of LIFT measurement in gel medium ................................................ 70
3.4.1 Monopolar ............................................................................................... 70

3.4.2 Bipolar ..................................................................................................... 71

 

 

 

3.4.3 Efl‘ect of Bubble Formation on LIFT Measurements .............................. 80
3.5 FERIT Measurements on the Monopolar RF Probe ....................................... 82
Chapter 4: Conclusions and Future Work 86
4.1 Future Developments for Application of LIFT and FERIT to RF Heating 89
4.2 Future Application for the Low Temperature FERIT Method ....................... 90
Appendix A: Procedure for Terbium and Europium Preparation 94
Appendix B: Light Source Evaluation with Ocean Optics Spectrometer ............... 97
Appendix C: Matlab Code for FERIT and LIFT Whole Field Analysis and AVI
Creation 100
BIBLIOGRAPHY 102

 

vi

LIST OF TABLES

Table 4.1: Comparison of conventional temperature measurement methods with
fluorescent temperature measurements ............................................. 88

vii

LIST OF FIGURES

Figure 1.1: Schematic of Bipolar and Monopolar Radiofrequency Heating .................... 4
Figure 1.2: Thermocouple Response in a Radiofiequency Field ..................................... 6
Figure 1.3: Experimental Configuration to Test the Presence of Water on IR
Measurements ........................................................................................................... 9
Figure 1.4: Effect of a Thin Fluid Layer on IR Thermometry ....................................... 10

Figure 1.5: Diagram of a dorsal skin flap chamber mounted on a rat, an actual skin flap
chamber, and measured isotherms by an IR camera resulting from low temperature

treatment of the tissue in the window chamber. ..................................................... 11
Figure 1.6: Fluorescence and Phosphorescence Diagram (Jablonski Diagram) ............. 13
Figure 1.7: Relationship Between Fluorescence Intensity and Quencher Concentration

at Various Temperatures ......................................................................................... 18
Figure 1.8: LIFT Experimental Schematic ..................................................................... 19
Figure 1.9: FERIT Experimental Schematic .................................................................. 21
Figure 2.1: Ocean Optics spectrometer sample holder with emission and excitation

fibers mounted 90° with respect to each other. ...................................................... 25
Figure 2.2: Light Source Evaluation with Spectrometer ................................................ 27
Figure 2.3: Radiofrequency probe coated with Europium (A), and plastic valve pintle

coated with Terbium (B) and excited with a 365nm UV light source. ................... 29
Figure 2.4: Terbium Emission Spectra for Excitation at 254nm and 365nm ................. 30
Figure 2.5: Fluorescent Emission Response of Europium and Terbium ........................ 31
Figure 2.6: Temperature Assignment for Measured Intensity Ratio .............................. 34
Figure 2.7: Comparison of Calibration Methods for the Europium Fluorophore .......... 35
Figure 2.8: Europium Temperature Dependent Fluorescence Intensity Profiles

Evaluated with a Digital Camera and Spectrometer ............................................... 37
Figure 2.9: Comparison of FERIT Measured Temperature with Thermocouple

Temperature ............................................................................................................ 39
Figure 2.10: Temperature Dependent Fluorescence Intensity Profile of Terbium ......... 41

viii

Figure 2.11: Schematic of the temperature control system used to create a cryogenic
flow environment. Not shown are the UV excitation source and digital camera. . 42

Figure 2.12: Comparison between thermocouple measurements and FERIT
measurements from 24°C to —80°C. The linear line was added to emphasize the
agreement between the data sets. ............................................................................ 43

Figure 2.13: Condensation Detection with the Terbium Polymer Film ......................... 44

Figure 2.14: Ability of FERIT to Detect the Onset of Condensation at Low
Temperatures .......................................................................................................... 45

Figure 2.15: Rhodamine B and Rhodamine 110 mixed in saline solution without
excitation illtnnination and with illumination by an Argon laser sheet (488nm) 46

Figure 2.16: Emission response of Rhodamine B at Various Excitation Frequencies 48
Figure 2.17: Emission Response of Rhodamine 110 at Various Excitation Wavelengths

................................................................................................................................ 49
Figure 2.18: Intensity variation of Rhodamine B, Rhodamine 110 solution .................. 52
Figure 2.19: Fluorescence Ratio of Rhodamine B and Rhodamine 110 with Respect to

Temperature ............................................................................................................ 53
Figure 2.20: Evaluation of the Light Source Used for Calibration of Rhodamine B and

Rhodamine 110 ....................................................................................................... 55
Figure 2.21: Spectral Emission Shift Due to Concentration of Rhodamine B ............... 57
Figure 2.22: Intensity Ratio Dependence on Temperature with Varying Concentrations

of Rhodamine B and 110 ........................................................................................ 58
Figure 3.1: Schematic of Validation Test ....................................................................... 60

Figure 3.2: Comparison of thermocouple measurements with LIFT temperature
measurements. Inset picture is a gray scale LIFT image showing the location of
the thermocouples and approximate location of the line profile. ........................... 62

Figure 3.3: Effect of Concentration of Rhodamine B and Rhodamine 110 on Emission
Intensity .................................................................................................................. 64

Figure 3.4: Image of monopolar and bipolar radiofrequency probes. Close up images
show the details of the probe tips. .......................................................................... 67

Figure 3.5: Image of monopolar RF Probe Submerged in an Optically Clear Gel
Medium ................................................................................................................... 68

ix

Figure 3.6: Experimental Equipment for LIFT measurements indicating the laser beam
path, RF probe submerged in an optically clear tissue phantom, RF generators, and

digital camera. ........................................................................................................ 69
Figure 3.7: LIFT Predicted Temperature Around the Monopolar RF Probe .................. 70
Figure 3.8: LIFT Predicted Temperature Around the Bipolar RF Probe ....................... 72

Figure 3.9: Gel phantom experimental configuration with the laser sheet passing
through the bottom of the glass test container. ....................................................... 73

Figure 3.10: Whole field temperature measurement in a gel and saline medium due to
monopolar RF heating. Line indicates gel-saline interface. Image A is l to 2
seconds post activation and image B is 45 seconds post activation. ...................... 76

Figure 3.11: Measured power output for the RF monopolar probe in a gel-saline
medium and in a saline solution only medium. ...................................................... 77

Figure 3.12: Whole field temperature measurement in a gel and saline medium due to
bipolar RF heating. Line indicates gel-saline interface. Image A is 1 to 2 seconds
post activation and image B is 100 seconds post activation. Arrows in image B

indicate bubbles. ..................................................................................................... 79
Figure 3.13: Bubble formation on RF monopolar probe tip. Bubbles create an optical

disturbance that effects LIFT measurement. .......................................................... 80
Figure 3.14: Temperature measurement effect of bubble formation. The points around

the probe tip represent the relative location of the measurements. ........................ 81
Figure 3.15: F BRIT Measurements on the Surface the Monopolar Probe Tip ............... 84

Figure 3.16: Comparison of FERIT Measurements with the Temperature Measured by
the Thermocouple Embedded in the Monopolar Probe Tip ................................... 85

Figure 4.1: Potential application of low temperature F ERIT to measure temperatures on
the window chamber during cryosurgery in order to correlate temperature and

thermal damage ....................................................................................................... 91
Figure 4.2: GFP Tumor Qualitative Intensity Measurements During Freeze-Thaw Cycle

................................................................................................................................ 92
Figure 4.3: Temperature dependent fluorescence of GFP fi'om —12°C to 60°C. ........... 93

Figure B 1: Evaluation of the Spectral Emission of Electromagnetic Excitation Sources
with the Ocean Optics Spectrometer ....................................................... 98

Images in this thesis are presented in color

c Speed of light

D Sum of the diffusion coefficients of fluorophore and quenchers
E Energy

fq Quenching efficiency

F Fluorescence intensity in the presence of a quencher
F0 Fluorescence intensity in the absence of a quencher
h Planck's constant

I Intensity

k Boltzman’s constant

k0 Diffusion controlled birnolecular rate

kq Birnolecular quenching constant

N Avogadros number

Q Concentration of the quencher

R Molecular Radius

T Temperature

x x-coordinate

y y-coordinate

Greek:

11 Solvent viscosity

3. Wavelength

v Wavenumber

Vf Frequency

to Lifetime

Subscripts:

em Emission

ex Excitation

EXP Experiment

REF Reference

Abbreviations:

FERIT Fluorescent Emission Ratiometric Imaging Thermography
GFP Green Fluorescent Protein

IR Infrared

LIFT Laser Induced Fluorescence Thermography
RF Radiofrequency

UV Ultraviolet

KEY TO SYMBOLS AND ABBREVIATIONS

Chapter 1: Introduction

Temperature measurement and acquisition is an indispensable requirement in
many areas of research and industry ranging from engineering and science to medicine.
There are many situations in these fields of study where an increasingly detailed
knowledge of temperature has become a necessity for developing and improving the
technologies of these industries. Some situations warrant the use of non-conventional
temperature measurement methods since conventional methods, such as thermocouples,
infrared cameras, and other commercially available temperature sensing equipment may
provide too little detail about temperature fields. They also may be unable to provide
data in harsh chemical or electromagnetic operating environments, and may be
geometrically incompatible with the application. This has led to the development of
new and innovative methods for measuring temperatures using luminescent chemicals.
Luminescence is the general term that encompasses the processes of fluorescence and
phosphorescence. The three terms are often used interchangeably. Fluorescence and
phosphorescence are similar in that they both generally result in the emission of visible
radiation commonly associated with black light posters and glow in the dark watch
faces. For the remainder of this thesis the terms fluorescence and luminescence will be

primarily used.

Fluorescent lighting and phosphor screens for televisions were some of the first

uses for luminescent chemicals. Developers of these technologies sought luminescent

chemicals that would not lose their intensity as temperatures increased. As early as
1933, Neubert suggested the use of luminous paints as an indicator of temperature.1 He
noticed that the intensity of luminescence of certain chemicals diminished at high
temperatures. This discovery has lead to the extensive study of the class of chemicals
that luminesce, i.e. fluorescent material, in order to understand their temperature
dependent characteristics. The common characteristic that allows these chemicals to be
used as a temperature indicator is the generally inverse relationship between

luminescent intensity and temperature.

This thesis describes the principles of fluorescence and the factors that affect
fluorescent intensity. Methods for calibration and quantification of fluorescent intensity
variation as an indicator of temperature are presented, and these methods are applied to
measure temperatures on the surface and in the surrounding medium of radio-fiequency
heating devices. Additional methods for whole field surface measurements down to
cryogenic temperatures are also explored. Two types of fluorescent thermography are
explored: Laser induced fluorescence thermography (LIFT), and fluorescent emission
ratiometric imaging thermography (F ERIT). Both techniques use the fluorescent
emission of specific chemicals as indicators of temperature. The LIFT technique is a
volumetric method performed in a fluid or solid medium whereas the FERIT method is
a surface temperature measurement method. The details of the techniques are described

later in the thesis.

1.1 Introduction to Radiofrequency Heating

Over the last decade, sub-ablative thermotherapy applied by laser and
radiofrequency heating has found ever-increasing application in orthopaedic and
ophthalmologic surgery. The applications are: the treatment of the instabilities of the

shoulder, ankle, knee and hip joints 2’3’4, elimination of discogenic pain in the spine 5'6,

0,11 12,13

tissue welding 7, thermokeratoplasty 8’9, chondroplasty l and skin resurfacing

Irrespective of the specific procedure applied, the aim of the therapy is the same:
altering the configuration of the strain-damaged, pathological or congenitally
anomalous soft tissue (whether it be skin, joint capsule, ligament, disc annulus, cartilage
or cornea) in order to re-establish stability and function. The basic mechanism of this
therapy is the heat-induced denaturation of the underlying collagenous network of the
soft tissues. Heat-induced denaturation causes micro-structural transformation of the
collagen molecule and produces a thermal stress field that is a function of the direction
of the collagen fiber orientation. The thermal stress field formed changes the
configuration of the tissue.

In the clinical setting, heat is applied by a laser or a radiofrequency (RF) probe
swept on the surface of the target tissue. An irrigation solution is circulated around the
target tissue in order to eliminate hot spots and to clean the debris. In laser heating, a
Ho:YAG laser is usually preferred. In radiofrequency heating however, there are two
alternatives. The first one is bipolar radiofrequency in which the active and the passive
electrodes are located at the probe tip. The second choice is monopolar radiofrequency

in which the active electrode is located at the tip of the probe and the return (patch)

electrode is placed on the patient’s skin. The two modes of radiofi'equency heating are

illustrated in Figure 1.1. This figure shows a schematic of the monopolar and bipolar
probes and shows the assumed general nature of the electrical field created between the

positive and negative return electrodes.

 

BIPOLAR MONOPOLAR
RADIOFREQUENCY RADIOFREQUENCY

    

r x ' '
130. '1Jo
Tissue

 

 

 

 

Figure 1.1: Schematic of Bipolar and Monopolar Radiofrequency Heating

Oratec Interventions supplied the monpolar radiofrequency unit and ArthroCare Inc.
supplied the bipolar radiofrequency unit. During the operation, the surgeon sweeps the
probe on the target until color change and shrinkage are observed. The success of the
therapy is primarily based on the visual perception and the technique of the surgeon.
Collagenous tissue denaturation and shrinkage in response to heating by laser,
microwave, radiofiequency devices and hydrotherrny are extensively studied
phenomena ”'15’16. The amount and the extent of the irreversibility of tissue shrinkage
(i.e., thermal-mechanical strain formation) depend on many factors. These factors

include the thermal history (temperature and time) ”"34”" as well as the mechanical

stress applied on the tissue during heating 2‘. Other factors include the pH of the
surrounding medium and electrolyte concentration 22 as well as the type of collagen 23,
its hydration level 24, and the degree of cross-linking 25. In other words, the amount and
rate of heat deposition, coupled to the properties of the tissue and the surrounding
medium govern the response at both the microscopic and macroscopic scales 26.

Creating accurate models and simulation tools that can predict the extent of
thermal damage penetration into the tissue when treated with different clinical heating
methods and examining the effects of clinical parameters (such as the amount of saline
circulation, variations in probe sweep speed, power setting and control algorithms) on
the outcome require extensive knowledge of the thermal fields generated by these
probes. This requires measurement of the temperature profiles (and therefore
calculation of the heat fluxes) created by these probes and temperature measurement of

the probe tip surfaces.

1.2 Evaluation of Common Temperature Measurement Techniques

1.2.] Thermocouples

Utilization of thermocouples in order to determine the temperature distribution
on and around the probes was not feasible due to the RF field generated by these
probes. Thermocouples are a convenient device for making accurate point
measurements with fast acquisition rates and. are well suited to measure temperatures at
the cryogenic level up to 3000°C 27. They are very easy to use and are an inexpensive
means for temperature measurement. However, thermocouples lack the ability to make

whole field measurements and potentially interfere with an experimental environment

such as a flow field or an in viva surgical procedure. Thermocouples are also limited to
the nature of the surrounding environment in which they operate. Corrosive chemicals
and radiofi'equency and microwave fields are known causes that result in faulty
operation of thermocouples. Figure 1.2 shows the effect a RF field has on a
thermocouple signal. A single uncalibrated thermocouple was placed within 5mm of
the bipolar radiofrequency probe tip for a set point of "1". Temperature was acquired at

1 Hz with the RF generator off and while the RF generator was activated.

 

120 .
3 RF on
A 100 J /
9 J
E 80 i \
a ;
E 60 7
8 i
,E, 40
l- . RF off RF off
20 a
o t__g__,i_ __ t T __*_
0 1o 20 30 4o 50
Time (s)

Figure 1.2: Thermocouple Response in a Radiofrequency Field

When the RF generator is turned on, the thermocouple signal becomes distorted and
produces faulty data. The thermocouple signal is restored after the generator is turned
off. The Oratec monopolar RF generator uses a T-type thermocouple embedded in the
probe tip as a reference for its power control settings. The thermocouple signal is

processed by a series of filters to reduce the noise produced by the radiofrequency

signal. The details of the filtering methods are unknown. It was observed in other
experiments that the bipolar generator had a greater effect on thermocouple signal than
the monopolar unit. This suggests it may be possible to filter thermocouple signals
depending on the strength of the radiofiequency field, but there might still be some
uncertainty in the temperature measurement depending upon the efficiency of the filter.
Since one of the aims of this thesis is to measure whole field temperatures around
radiofrequency heating devices, thermocouples were not deemed a viable means for

temperature measurement.

1.2.2 Fluoroptic Thermometry

Fluoroptic thermometry was not feasible as well due to the relatively large size
of the probes in the presence of large gradients. F luoroptic thermometry is the optical
analog to a thermocouple. It is similar to thermocouple measurements in that it
measures temperature at a "point". F luoroptic thermometry uses a temperature sensitive
phosphor that has been applied to the end of a quartz optical fiber. Light pulses, which
excite the phosphor, are launched down the optical fiber by a control unit and the
resulting fluorescence decay response after each pulse is measured using the same
optical fiber. The measured decay response varies according to the temperature of the
tip of the optical fiber. F luoroptic thermometry systems can measure temperatures from
-200°C to 450°C and are capable of temperature measurements in harsh environments
such as radiofi'equency and electromagnetic fields.28 However, fluoroptic thermometry
is limited by its relatively slow sample rate of 4H2 and a relatively large fiber optic
diameter of 1mm. The work described in this thesis uses similar fluorescent methods

for whole field temperature measurement.

1.2.3 Infrared Thermography

Another alternative was using an Infiared (IR) camera. However, light at this
wavelength is absorbed by the saline (both probes should be immersed in saline to
operate) over very short distances and it was not possible to collect any information
below the surface. IR thermometry equipment can provide whole field temperature
measurement at flaming rates up to or greater than 33Hz and is relatively simple to use
since it is essentially a "point and shoot" method. They are often used in a
preventative engineering fashion to monitor hotspots in applications ranging from
industrial equipment to electrical circuit boards. IR can measure temperatures from -
40°C up to 2000°C.29 Due to the nature of thermal radiation; an electromagnetic wave
traveling through a medium may become attenuated. If attenuation is such that no
radiation emerges on the other side of the medium, the medium is known as opaque.30
This property is also dependent on the thickness of the medium. Infrared radiation does
not pass through media such as water and glass. This is unlike the visible spectrum,
which readily passes through these media for considerable distances before any
absorption takes place. This characteristic of infrared radiation makes it impossible to
use an infiared camera to measure the temperature of an object submerged in water.
The RF probes studied in this thesis are operated in saline solution at depths greater
than 5mm from the surface. This depth is too great for infrared radiation to penetrate so
the camera reports the surface temperature of the saline and cannot acquire temperature

measurements around the RF probe.

A test was performed to determine the effect a thin layer of water covering a

heated surface has on IR temperature measurements. The bipolar radiofrequency probe

was used to heat the underside of a bologna sample and temperatme was measured on
the opposite side with the TR camera and fluoroptic thermometer. This was initially
done to gage the performance of the probes and to become familiar with their operation.
The fluoroptic thermometer was in contact with the top surface of the bologna sample
and the RF probe was in contact with the underside of the sample directly below the
fluoroptic fiber as shown in Figure 1.3. The probe was activated and temperature
measured by the IR camera near the fluoroptic fiber was compared to the measured
temperature of the fluoroptic thermometer. A layer of water approximately 1mm thick

was added to the surface and the same measurements were performed.

 

 

F luoroptic fiber \T l
nfrared camera

Water on surface

 

Bologna sample

ID

Saline

 

 

 

 

\ArthroCare bi-polar probe

Figure 1.3: Experimental Configuration to Test the Presence of Water on IR
Measurements

Figure 1.4 illustrates the effect the layer of water had on temperature measurement with
an infiared camera when compared to measurements from a fluoroptic thermometer.

The probe was activated for approximately 11s for both the wet and dry situations. The

results indicated water had the effect of damping the maximum temperature measured
by the camera. For the dry situation, the IR camera and the fluoroptic thermometer
agreed within 1°C of each other for the transient temperature cycle. When a thin layer
of water was added to the surface the fluoroptic and IR camera differed by more than
10°C at the time of the maximum observed temperatures.

70 t
I Fluoroptic no liquid

60 1| IR no quuid\ Fluoroptic liquid

0|
0

i
l

.5
O

(A)
O

i

 

Temperature (°C)
N
O

.5
o

l
0 20 4O 60 80 100

Figure 1.4: Effect of a Thin Fluid Layer on IR Thermometry

This makes it impossible to measure transient temperatures on a surface submerged in a
fluid due to the thermal damping effects of the water and the opaque nature of water for

relatively thin layers of water.

Therefore, in this thesis, an attempt has been made to utilize LIFT and FERIT as a
method for measuring temperature fields in the surrounding medium and on the surface

of monopolar and bipolar radiofiequency probes.

10

1.3 Introduction to Temperature Measurement at Cryogenic Levels

Whole field temperature measurement at cryogenic levels (-80°C) is desired since
commercially available methods to make whole field measurements down to —80°C
does not exist. Infrared cameras are limited to a minimum temperature of —20°C to —
40°C depending upon the sensitivity of the 1R detector. Thermocouples are capable of
measurement down to cryogenic levels but can only provide point measurements and
introduce fin effects especially for large thermal differences between the thermocouple
tip and the ambient environment. Low temperature whole field measurements have
utility in applications such as cryomicroscopy and the characterization of cryosurgical
probes in tissue phantoms during freeze-thaw cycles. Figure 1.5 illustrates an example
where whole field temperature measurement would be beneficial. Figure 1.5A is a
diagram of a rat fitted with a dorsal skin flap chamber. The chamber as shown in Figure
1.58 is approximately 1 cm in diameter and is used for studying low temperature

thermal treatments of tumors.

 

Figure 1.5: Diagram of a dorsal skin flap chamber mounted on a rat, an actual
skin flap chamber, and measured isotherms by an IR camera resulting from low
temperature treatment of the tissue in the window chamber.

Cryosurgery is performed on the tissue in the chamber and detailed temperature
information is necessary for complete understanding of the resulting tissue damage.
Figure 1.5C is a picture of isotherms in the window chamber resulting fiom treatment
by a cryoprobe measured by an infrared camera The minimum temperatures recorded
were near -20°C, but lower temperatures were evident since the IR detector became
saturated for regions out of its measurement range. However, new methods are needed
to make the same type of measurements down to ~80°C. Therefore fluorescent
temperature measurement techniques have been developed to make it possible to

perform whole field measurements at cryogenic temperature levels.

1.4 Fundamentals of Luminescence

Luminescence is the term that encompasses fluorescence and phosphorescence
and they are often used interchangeably even though there are specific differences.
Luminescence is generally described as the emission of light from any substance that

l A substance is caused to luminesce by an

occurs from electronically excited states. 3
influx of energy to the substance, which causes the electrons to become excited to a
higher electronic state. This can be accomplished by exposure to electromagnetic
radiation including visible and ultraviolet light, x rays and gamma rays, particle beams
of electrons, neutrons, or ions, and electrical currents.32 Luminescence is formally
divided into two categories, fluorescence and phosphorescence. Fluorescence refers to
light emission that is usually in the visible spectral band and has a duration of 10'9 to 10'
3

8, whereas phosphorescence is associated with a longer duration, typically 10'3 —- 103 s.

A common test to determine if a material is fluorescent or phosphorescent is to simply

12

excite the material with the excitation source. If the material continues to luminesce
after the excitation source is turned off, the material is phosphorescent. The material is
fluorescent if the luminescence decays rapidly or immediately after the excitation
source is turned off. Figure 1.6 is an energy level diagram or Jablonski diagram and is a
physical representation of fluorescence and phosphorescence. Before excitation, the
electrons in a material are populated in the ground state. In order for excitation to
occur, a means to deposit energy (a photon) is needed and is most often in some other
form of electromagnetic radiation such as visible or ultraviolet light. When a photon is
absorbed, the electrons transition from the ground state to an excited state of the same

spin multiplicity (81 or 82).33

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Singlet excited states Triplet excited state
1 Internal Vibrational
conversion relaxation
82 ‘ / Intersystem
J ‘ ¥ \ r crossing
ll " ll 1
s. t
v: ‘ T}
I
I l l I I
g I I I I" I
g I I I I I"
“' Internal l l l l I
A . 1‘1 and Phl I l I I
bsorptron uorescence external losrlrbolres'eere
conversion I I I I I
I I I I I
I I I I I
I I I I I
1 I I I I i
Vibrational :- ; I I
Ground 1 D i
am '12 Al Iv'r Ar 13 14

Figure 1.6: Fluorescence and Phosphorescence Diagram (Jablonski Diagram)

13

The excited electrons now undergo vibrational relaxation as a result of the vibrations in
the molecule and collisions with surrounding molecules and solvent. The electrons
relax to the ground vibrational level in their respective energy level they are located in
with out any fluorescent emission. This takes approximately 10425. Internal conversion
follows next where the electrons transition between electronic states (82 to 81). This is
dependent on whether the electron was initially excited to the S] or S2 state. From the
ground vibrational level of the S] excitation state, the electron can return to the ground
state (SO) by emitting a photon. This is known as fluorescent emission and can only
occur due to transitions from the 81 state to the SO state. The excited electrons in the
81 state can also undergo spin conversion resulting in the formation of a triplet state,
Tl. This is described as intersystem crossing. Transitions from the T1 to the SO state
are forbidden due to the mismatch in electron spin. This is why this transition takes
much longer than a transition fiom $1 to SO. This phenomenon is known as
phosphorescence and is the emission of light from a triplet-excited state. Emission rates
are slow for this phenomenon and can take as long as several minutes as the excited
phosphors return to the ground state. Whether radiative emission occurs from the 81
state or the T1 state the electrons undergo external conversion, which involves energy
transfer or collisions between the excited fluorophore and the surrounding solvent or
matrix. This serves to reduce the radiated intensity of the emitted photon. The
J ablonski diagram in Figure 1.6 shows that the energy of emission is typically less than
that of absorption so fluorescence and phosphorescence occur at lower energy levels or

longer wavelengths than absorption wavelengths. This phenomenon is known as Stokes

l4

shift since it was first observed by Sir G. G. Stokes in 1852. This phenomenon is also

represented by the Einstein photoelectric law

he
E=hvf=hcv=7 (1.1)

where E is the energy, his Planck’s constant, Vf is the frequency, c is the speed of light,
v is the wavenurnber, and I. is the wavelength”. Since h and c are constant, B and 2. are
inversely related to each other. A reduction in energy results in an increase in
wavelength. The energy absorbed by an electron is not equal to that released by the
electron since some energy is lost due to vibrations and collisions mentioned
previously.

Fluorescence and phosphorescence emission have been found to be dependent
upon certain environmental variables such as pH, pressure, temperature, oxygen
concentration, etc.3 1’32’35’36 These variables can be related to events such as the intensity
of the radiated emission, the lifetime of the emission, and the spectral shift of the
emission. The most common and convenient emission parameter used to measure
variables such as the pH, pressure, and temperature is the intensity of emission. The
work presented in this thesis focuses on the use of fluorescent chemicals to measure
whole field temperature distributions by associating measured fluorescent intensity

changes to changes in temperature.

15

1.5 Theory of Temperature Dependent Fluorescence

Several factors result in the decay or degradation of fluorescence intensity. These
factors are known formally as quenchers of fluorescence. Quenching does not involve a
permanent change in the fluorescent molecule rather it is a temporary degradation of
fluorescence until the quencher diminishes. Quenchers can take the form of a physical
material introduced into the environment or can be a phenomenological change such as
a temperature or pH change. The addition of a foreign material to the environment of
the fluorescent chemical increases the random collisional encounters between
fluorophore and quencher. A temperature increase causes the fluorescent molecules to
experience increased collisions with the solvent or matrix molecules. In either case,
random collisional encounters between fluorophore and the quencher or environment
are increased and this reduces the observable fluorescence emission. This unique
characteristic of fluorescent chemicals enables them to be used to detect trace amounts
of foreign material in a fluid in addition to thermographic measurement applications.
This thesis does not focus on the detecting foreign particles in a solution but rather on
the phenomena of temperature dependent fluorescence. The effect of collisional

quenching can be described by the Stern-Volmer equation,

F
F0 = 1+ quo [Q] (11)

In this equation, F0 is the fluorescence intensity in the absence of the quencher, F is the
fluorescence intensity in the presence of the quencher, kq is the bimolecular quenching

constant, to is the lifetime of the fluorophore in the absence of the quencher, and Q is

16

the concentration of the quencher3 1. The bimolecular quenching constant, kq, is

dependent upon temperature via the following three equations.

kq = kao (13)

In this equation, k0 is the diffirsion controlled bimolecular rate and fQ is the quenching
efficiency. The term, fQ, is a constant value and reflects how effective a collisional

encounter is at quenching a fluorophore. The term k0 is a constant given by,

k _ 47rRDN
0 1000 “'4’

where R is the collision radius, D is the sum of the diffusion coefficients of the

fluorophore and quencher, and N is Avogadro’s number. The diffusion coefficient, D,

is given by,
[CT
D = —
672771? (1.5)

where k is the Boltzrnan constant, 1] is the solvent viscosity, R is the molecular radius
and T is the temperature. Equation 1.5 indicates that the diffusion coefficient increases
with increasing temperature, which results in an increase in kg in Equation 1.4, thus
increasing the bimolecular quenching constant, kq, in Equation 1.3. This causes the
ratio F o/F in Equation 1.2 to increase thus denoting a decrease in intensity with
temperature increase. Figure 1.7 is a plot of Fo/F versus [Q] with a slope equal to quo
for arbitrary scenarios, This plot illustrates a linear relationship between the quencher

concentrations to fluorescence intensity. If the quencher concentration [Q] remains

17

constant throughout an experiment and temperature is increased, then temperature is the

only parameter causing a decrease or increase in the fluorescence intensity value F.

 

   

 

 

1 4 T2
j T1
A

1.1.
E - r1<rz<rs

‘ Q=constant

1 a
[0]

Figure 1.7: Relationship Between Fluorescence Intensity and Quencher
Concentration at Various Temperatures

For the experiments in the following chapters the concentration of the quenchers
remained constant therefore allowing direct correlations to be made between

fluorescence and temperature for the LIFT and FERIT methods described next.

1.6 Introduction to LIFT

Laser induced fluorescence thermography is a temperature measurement method
that uses a laser sheet to excite fluorescent chemicals in a solution or matrix. A laser is
used since it can be converted into a sheet in order to excite a plane in the fluid medium.
Incandescent light sources can also be used to excite the solution, but yield a volumetric
average of fluorescent intensity and therefore a volumetric average of the temperature.

An incandescent excitation source was used for calibration procedures presented later

18

when the fluid medium was maintained at isothermal temperatures. A sensitive
detector, usually a high quantum efficiency digital camera or spectrometer, measures
the fluorescent emission. The measured emission intensity is then converted to
temperature based on calibration curves. Figure 1.8 is a schematic of the LIFT

temperature measurement technique.

Heat source

    
  
   

Laser sheet excites a planar
region of interest in the fluid

Argon laser (lax) \
/'

Cylindrical Lens

Liquid solution with
dissolved fluorescent
chemical

 

S Xem>xex

_ Barrier filter

4a.

Figure 1.8: LIFT Experimental Schematic

High QE digital
camera

   

A laser beam is transformed into a laser sheet with a cylindrical lens. The laser sheet
1mm thick excites a planar area in the fluid medium containing fluorescent chemicals
that are excited by the wavelength emitted by the laser. For an incandescent light
source, a barrier filter is used to separate the particular wavelength of light that excites
the fluorescent chemicals in solution. The laser sheet can be positioned in any plane of

interest but is illustrated as impinging the centerline of the submerged heat source so

that planar fluorescence intensity around the submerged heat source can be measured.
The resulting fluorescent emission is separated from all other wavelengths of
electromagnetic radiation with a barrier filter and the intensity of the emission is
measured by a sensitive digital camera or spectrometer. The resulting fluorescence
emission intensity is much less than that of the excitation source; so sensitive emission
detectors are necessary. A digital camera produces a gray scale image as shown in
Figure 1.8 and temperature is extracted at all points of illumination using calibration
curve fit analysis. This method will be demonstrated for temperature measurement in a
fluid medium surrounding the radiofrequency probes as well as within a clear tissue

phantom.

1.7 Introduction to FERIT

Fluorescent emission ratiometric imaging thermography is a whole field surface
temperature measurement method that uses fluorescent chemicals embedded in a
polymer film coating as is illustrated in Figure 1.9. The fluorescent film can be coated
on most non-permeable surfaces and is excited with an incandescent source with a
barrier filter to select the desired excitation wavelength. The fluorescent emission
intensity fi'om the surface of the sample is measured with a sensitive detector with a
barrier filter and converted to temperature in a similar manner as the LIFT method. The
FERIT technique is suitable to measure temperatures for cryogenic applications as well

as the high temperatures associated with radiofrequency heating.

20

Sensitive fluorescence

detector with barrier filter Fluorescent sample
W\/\ km, if

Fluorescence l'

emission from sample

     

 

vat 7a.

I

_ Barrier filter

\\//

 

' Excitation source
Fluorescence intensity I pr “1‘11““3 (X) nm
image converted to wave ength
temperature data

Figure 1.9: FERIT Experimental Schematic

This method will be demonstrated later on for surface temperature measurements on

objects cooled to -80°C as well as for the radiofiequency probes heated to 85°C.

21

Chapter 2: Evaluation of Temperature Dependent Fluorescent
Chemicals

The materials and procedures required for a fluorescent dependent temperature
measurement system are relatively straightforward. The basic equipment needed is an
electromagnetic energy source that causes the temperature dependent fluorescent
chemical to be excited, a sensitive detector used to measure the resulting fluorescent
emission and data acquisition equipment to record the detector output. Also needed is a
means to calibrate the fluorescent intensity emission of the fluorophore with
temperature. Many variables can affect temperature measurement with fluorescent
chemicals so a complete understanding of the response of the fluorescent chemical is
needed for the environment in which testing is to be performed. The first step in using
fluorescent chemicals, as an indicator of temperature, is to select a chemical from the
list of many known fluorescent chemicals, and to establish or verify that the chemical
exhibits temperature dependent fluorescence within the temperature range of the
specific application. There are many fluorescent chemicals that have been evaluated by
other researchers so a literature search is an essential resource for chemical selection.
The chemical must also be able to co-exist in the operating environment of the
application. For example, some fluorescent chemicals are quenched by a pH change, so
using a chemical in an environment with fluctuating pH and temperature may result in

faulty measurements since pH and temperature would both affect fluorescence intensity

22

simultaneously. Based on previous research done in this lab and on a literature search,
the fluorescent chemicals selected for evaluation of the FERIT method are Europium
(1H) Chloride, and Terbium (IH) Chloride. See Appendix A for procedures regarding
the chemical preparations of these compounds. Both Europium and Terbium exhibit
strong temperature dependent fluorescence and are used individually for specific
temperature ranges. Temperature is obtained with ratiometric methods described in
detail in the next sections.

The chemicals selected for the LIFT method are Rhodamine B and Rhodamine
110. These chemicals are used in combination for a single temperature range.
Rhodamine B exhibits strong temperature dependence and Rhodamine 110 is relatively
temperature invariant. Rhodamine B and Rhodamine 110 fluoresce at different
wavelengths and it will be shown later that the ratio between the respective fluorescence
intensities produces temperature measurements that are presumably invariant with
respect to experimental variables such as excitation source fluctuation and other optical
aberrations. The Europium and Terbium are utilized in a polymer that is applied
directly to a surface of interest by an airbrush, whereas the Rhodarnines are utilized in a
fluid as a volumetric indicator of temperature. The temperature measurement
procedures for the liquid based and solid based fluorescent materials vary, so they will

be divided into these sub categories for the proceeding discussions.

23

2.1 Evaluation of Fluorophores with a Spectrometer

A spectrometer is an instrument that measures electromagnetic radiation
intensities for a range of wavelengths. Evaluation of the fluorophores with a
spectrometer is essential in determining the optimum excitation wavelength of the
fluorophores and for proper selection of barrier filters used for measurement of the
emission intensities. The sensitivity of the spectrometer to particular wavelengths is
dependent on specifications of the detector. The model used in this research, an S2000
spectrometer (Ocean Optics, Dunedin, FL) was capable of measuring wavelength
intensities from low UV (250nm) to near IR (800nm). Wavelengths beyond these
ranges were not encountered so spectrometer performance is unknown outside of this
wavelength range. Some spectrometers are also capable of controlling the excitation
wavelengths with lOnm or better bandpass control fi'om UV to IR wavelengths. The
Ocean Optics spectrometer came equipped with a dual excitation light source that
allowed for UV and/or visible broadband excitation wavelengths. Specific excitation
wavelengths could be achieved with barrier filters. Figure 2.1 shows an image of the
spectrometer sample holder which is located at the center of the image and is sized to
hold a 1cm2 cuvette for fluid analysis or solid fluorescent samples that are coated on a

substrate of comparable size.

24

Emission detector fiber

re. ‘

Excitation source fiber

Heating and cooling
circulation bath inlet/outlet

 

Figure 2.1: Ocean Optics spectrometer sample holder with emission and excitation
fibers mounted 90° with respect to each other.

The sample holder is mounted onto a hollow aluminum base that allows the fluorescent
samples to be temperature controlled by connecting a constant temperature circulation
bath using the hoses at the right of the image. The circulation bath was capable of
controlling the temperature of the specimens fi'om —12°C to 100°C. The spectrometer
sample holder can withstand temperatures fi'om -100°C to +100°C. The manufacturer
did not recommend temperatures outside of this range. A fiber optic cable directs light
from an excitation source into the sample holder and the fiber optic 90° clockwise fiom
the excitation fiber optic directs all emitted wavelengths from the fluorescent sample to
the spectrometer detector for signal processing.

In addition to evaluating fluorescent specimens, the spectrometer is useful for
evaluation of barrier filters and excitation light sources. A barrier filter can be placed in

the slot denoted in Figure 2.1 in hem of the emission detector with an excitation source

25

that encompasses the wavelengths of the barrier filter pointed directly at it. The
excitation fiber can be moved 180° from the emission fiber to direct light at the
emission fiber or any light source can be directed at the emission fiber. The
spectrometer should only register the wavelengths the barrier filter transmits. Thus the
manufacturer’s specifications of the particular barrier filter can be verified or
discredited.

Excitation sources can be evaluated in the same manner except the barrier filter
should be replaced with neutral density filters to avoid saturation of the spectrometer
detector. While evaluating the light sources, up to three neutral density filters with
2.5%, 4.0%, and 20% transmission were used at one time for an intense excitation
source. This filtering resulted in spectrometer integration times on the order of lOms.
The dual excitation light source equipped with the spectrometer was evaluated with its
own the detector to determine the specific wavelengths produced by the source in order
to verify manufacturer specifications. This was done because during the process of
evaluating some of the fluorescent samples, additional emission peaks were sometimes
observed dependent upon the particular excitation source being used. Since the location
and magnitude of the emission peaks are of critical importance in evaluating the
temperature dependence of a fluorescent chemical, a thorough evaluation of the
excitation sources was performed and the comprehensive results can be found in
Appendix B. The light source supplied with the Ocean Optics spectrometer is equipped
with Deuterium and Tungsten bulbs. The wavelength specification listed for the
Deuterium bulb was 200 to 400nm and the Tungsten bulb was 360 to 850nm. Figure

2.2 shows the results of the light source evaluation with the Ocean Optics spectrometer.

26

Intensity/l»...x

 

 

 

Wavelength (nm)

---»-~v---~--vTungsten bulb~— Deuterium bulb

Figure 2.2: Light Source Evaluation with Spectrometer

Light sources were evaluated by lining up the detector fiber cable with the light source.
Neutral density filters were placed in between the detector and light source as needed to
prevent the detector from becoming saturated. Figure 2.2 indicates that the Deuterium
bulb supplied with the spectrometer produced wavelengths that extend beyond the UV
range. This particular source was initially being used to evaluate the Terbium and
Europium compounds since those specimens are excited by ultraviolet radiation. It was
assumed that the Deuterium excitation source produced wavelengths in the UV region
only. During initial evaluation of Terbium and Europium with the spectrometer,
emission peaks at wavelengths different fiom published spectra data were recorded by
the spectrometer in the visible spectra region. These emission peaks were thought to be

fluorescent emission from the sample, however they were a result of the excitation

27

source. This illustrates the importance of characterizing the excitation source so that the

fluorescence experiments will not contain emission artifact.

2.2 Evaluation of the FERIT Fluorescent Chemicals

As indicated previously, the Europium and Terbium have been utilized in solid
form as a surface coating or “fluorescent paint”. The Europium and Terbium
compounds are chemically prepared in the lab by procedures indicated in Appendix A
and are dissolved in an organic solution containing dissolved PMMA pellets. While
still in solution the compound is airbrushed onto the object being tested. The solution
dries leaving behind a thin, polymer coating with the Europimn or Terbium embedded
in the polymer. The film is applied to the object until a white frosting becomes evident
on the surface. The object is then gently heated to approximately 70°C for a couple of
minutes to "cure" the film on the surface. The Europium compound has been
previously developed in the lab by Lian37 and exhibits temperature dependence from
20°C to 90°C. The temperature range of the Europium compound was limited by the
temperature range of the polymer film. The polymer film melts at temperatures greater
than 90°C. It has been suggested that Europium can be used to measure temperatures
up to 900°C.38 The method has been extended to cryogenic temperatures by using
Terbium, which has been shown to exhibit temperature dependent fluorescence down to
470°C.” The research presented here will validate the temperature dependence of
Terbium down to -80°C and apply the Europium compound for measurement on
radiofrequency probe tips. Figure 2.3A shows a radiofrequency probe coated with the

Europium polymer film and Figure 2.3B shows a plastic valve pintle coated with

28

Terbium. The specific use of the pintle will not be discussed in this thesis due to non-
disclosure issues but is presented for validation purposes of the temperature

measurement method only.

4,:

l
I
z

Europium (24°C to 90°C) Terbium (-80°C to 240)

 

Figure 2.3: Radiofrequency probe coated with Europium (A), and plastic valve
pintle coated with Terbium (B) and excited with a 365nm UV light source.

The samples are illuminated by the Opti Quip ultraviolet source with a 20nm bandpass

filter centered at 365nm, and are imaged with a handheld Sony digital camera.

2.3 Spectral Evaluation of the Polymer Based Fluorophores with the Ocean
Optics Spectrometer

Europium and Terbium are evaluated together since they are made in similar
fashion, but they are different in their fluorescent response to temperature. Both
fluorophores were evaluated with the Ocean Optics spectrometer by airbrush coating
two pieces of aluminum 1.4cm x 1.4cm with the Terbium and Europium polymer
compounds respectively. The coated sample was placed at a 45° angle in the sample
holder. It has been suggested that the Terbium and Europium are excited by ultraviolet

radiation.32 Previous work with Europium performed in this lab by Lian37 used the Opti

29

Quip UV source centered at 365nm and other cited research has suggested optimum UV
excitation wavelengths ranging from 254nm to 378nm.40 A study was performed to
determine the effect the UV excitation wavelength has on the Europium and Terbium
emission spectra. A new excitation source (Sirchie) that produced maximum intensity
at 254nm, with smaller peaks located at 315nm, and 365 nm was tested against the
original UV source that produced radiation at only 365nm with a 20nm bandwidth.
This was done to determine if the emission output of Europium and Terbium could be
improved by the higher energy excitation wavelengths included in the new source. The
spectral output of the two light sources can be found in Appendix (B). Figure 2.4 is the
resulting emission spectra of Terbium associated with excitation by the Sirchie (254nm)

and Opti Quip (365nm) UV sources.

 

 

Intensity (arbitrary units)

i

 

450 500 550 6(1) 650

Wavelength (nm)
--——Tb 254nm —— Tb 365nm

Figure 2.4: Terbium Emission Spectra for Excitation at 254nm and 365nm

The data sets have been normalized with respect to their own maximum intensities and

have been shifted on the y-axis in order to distinguish between the two sets. The figure

30

shows that the emission spectrum of Terbium has the same emission peaks for either
UV source, however the overall magnitude of emission intensity is dependent on the
power of the particular excitation source and the integration time of the detector. The
main emission peak of Terbium is located at 547nm or the green portion of the visible
spectrum. Europium was evaluated next and Figure 2.5 shows the emission spectrum of
Europium resulting from the excitation of the two UV sources. Europium has its main
peak at 615nm, the red portion of the visible spectrum, with an additional emission peak
at 595nm and ftuther emission at wavelengths greater than 650nm. Optical bandpass
filters were selected based on both emission plots so that fluorescence intensity fi'om

only the main emission peaks would be measured.

 

 

Intensity (arbitrary units)

 

Wavelength (nm)

Eu 254nm Eu 365nm

 

 

Figure 2.5: Fluorescent Emission Response of Europium and Terbium

A key point, in regards to the two UV light sources evaluated with the Terbium and

Europium fluorophores, is that the 254nm somce cannot excite a fluorescent sample

31

behind glass or submerged in water, whereas the 365nm source has the ability to excite
the fluorophores through water greater than 10cm thick. Glass and water do not
transmit the short UV wavelengths, therefore creating an obstacle for applications
where the excitation radiation must pass through these materials. For this reason, the
365nm UV source will be used for the excitation of Terbium and Europium samples in

future experiments.

The polymer-based fluorescent chemicals can be applied to most hard surfaces
such as glass, metals, and plastics with an airbrush, which produces a thin, fluorescent
film on the test specimen. The chemicals do not adhere well to porous and pliable
surfaces such as biological material. The coated surfaces are relatively durable and are
resistant to water and temperatures from ~196°C to 90°C. The polymer films have been
successfully submerged in liquid nitrogen, but they begin to melt at temperatures above
90°C. The fluorescent compounds of Terbium and Europium have been embedded in
other materials, which have greater temperature resistant capabilities.41 During the
application of the fluorescent film, it is difficult or impossible to ensure even
distribution or concentration of the fluorophore on the surface of the test object. This
uneven distribution produces noticeable variations in fluorescence emission intensities
for isothermal conditions. See Figure 2.3 for an image where the polymer film coating
is uneven. The methods described in the next section show how non-uniform
fluorophore concentration and other optical anomalies are eliminated by ratiometric

methods.

32

2.4 FERIT Ratiometric Methods for the Determination of Temperature

The F BRIT method measures fluorescent intensity and converts measured
intensity changes into temperatures. The ratiometric approach to the thermography
method ensures that the measured fluorescent intensity changes are due to temperature
changes only and not other experimental variables such as excitation source fluctuation,
non-uniform local concentration of the fluorophore, optical aberrations such as shadows
or dust particles, and the solid angle involved when imaging a curved surface.

For an experiment occurring between Tmin and Tm arbitrary temperatures, the
following procedures are followed. The specimen is maintained at an isothermal
starting point, usually room temperature, and verified with calibrated thermocouples.
The sample is then illuminated with the excitation source and imaged with the digital
camera or the intensity is measured with the spectrometer. The intensities in the image
are denoted Imax if the Europium compound is used and is denoted 1min if the Terbium
compound is used. This is because Europium has its maximum intensity at room
temperature and Terbium has its maximum intensity at —80°C. The actual test then
proceeds where the sample is cycled within the temperature range of the particular
fluorophore being used and imaged at pre-determined time intervals dependent upon the
transient nature of the experiment. After the test, an isothermal image at the other
extreme of the temperature range needs to be taken. This requires isothermal control of
the experimental system and is often the most difficult data point to implement in an
experiment. For example, a sample coated with Europium would be held at a
temperature higher than the temperatures in the experiment. The Tm“, (1min), endpoint

usually ranges from 80°C to 90°C for Europium. Once the endpoint image is taken,

33

temperature can be extracted from all images by performing the following operation at

each pixel:

( ’ ) mm( s )
%”—1"" = Irma.» (2.1)

max(i,j) — mm(i,j)

The intensity ratio is then compared to the intensities in the calibration data and

assigned a temperature value as illustrated in Figure 2.6.

Tealib Icalib

    

Figure 2.6: Temperature Assignment for Measured Intensity Ratio

This method measures the magnitude of the overall intensity change fiom Tmin to Tule
and assigns a temperature based on the percent change in intensity between the
maximum and minimum intensities. For an arbitrary situation of two independent tests
conducted over the same temperature range, it is possible to have one test experience an
overall magnitude change of 100 and the other test yield an overall change of 1000.
Corresponding temperatures for each test will have different absolute intensities but
have the same intensity ratio per Equation 2.1.

Other ratiometric techniques were also performed with the FERIT method that

made experiments easier to conduct. These involved only the initial isothermal

34

reference image to extract temperatures. The intensity for the initial isothermal
condition is imaged and used as the only reference frame for the entire experiment.
Equation 2.2 is used to create an intensity ratio. From this, a temperature is assigned

from calibration data in the same fashion as the first ratiometric method.

I

image(i,j) = I

I ratio (2.2)

ref(i,j)

The two ratiometric calibration methods are compared against each other in Figure 2.7

for a sample size of n=4.

—5

9.0.0
Moore

0

.o . .
U ¥ 0" Q
-_._ _J_._ _. __._..:__ _L_ _._ .__-_r._ _r_ .- ..
l.

Normalized Intensity
p o

.o .o

-I N

__ l_____ .‘__.
I

3
l
l
I
l

"7 ‘- T" W " —“'_’T " "— ‘_W* ‘ T— __'— "A —" —‘..

48 56 64 72 80
Temperature ('0)

5‘:
g .
8

o lllmax I (l—htin)/(inax-Inin)

Figure 2.7: Comparison of Calibration Methods for the Europium Fluorophore

35

The two methods produce different curves but the data show that both methods are
reproducible. The first ratiometric method (Equation 2.1) proved to be universal for
any specimen being tested. This will be shown in Figure 2.8 where multiple samples
were tested and shown to produce the same calibration curve. The second method
(Equation 2.2) however does not exhibit reproducibility from sample to sample. This
means that a specimen coated with the fluorophore must be calibrated prior to the actual
test. The calibration is specific for the test specimen. On the positive side, a second
reference intensity image at the opposite temperature extreme does not need to be taken

with this method, thus eliminating the need for additional temperature controls.

2.5 Temperature Dependent Fluorescence Evaluation of Europium

The temperature dependent fluorescence of Europium was measured previously
in this laboratory with the Spectra Source digital camera and barrier filter by Lian”.
The same sample was evaluated three years later by similar methods to acquaint the
author with the thermography technique and to evaluate the durability of the fluorescent
film. The sample consisted of a 1mm thick, 5cm x 5cm, aluminum sheet that was
coated with the Europium polymer compound attached to an electric foil heater of the

same dimension.

An experimental configuration similar to Figure 1.9 was used. Using a heater,
the temperature of the sample was increased in steps between room temperature, 24°C,
and 80°C at approximately ten-degree increments. A calibrated foil thermocouple

attached to the aluminum sample provided the independent reference temperature

36

measurement. Fluorescence intensity was measured in gray scale format by a sensitive
digital camera when the sample reached equilibrium for each set point temperature.
The Europium sample prepared for the spectral evaluation section of this chapter was
also cycled fiom 24°C to 80°C with the temperature control of the spectrometer. A
calibrated thermocouple was attached to the aluminum substrate and fluorescent
emission was recorded at 10°C increments up to 80°C. Figure 2.8 compares the
temperature dependent fluorescence intensity data obtained by the original researcher
with the data obtained for the same specimen and experimental methods by the author.
The data obtained for a new specimen using the spectrometer instead of the digital

camera is also shown.

0.8 «E '

I
0.7 4

.0
a:

0.5 l

.o .o
w #u
__.L.

Normalized Intensity

0.2 —J,

.0
o A
r. __L_e _

 

——— .~-— —— a —— -— —e

32 40 4e 56 64 72 80
Temperature (“Cl

N
.h

0 Digital Camera data (n=4) - Spectrometer data (n=1) A Previous calibration (n=3)

Figure 2.8: Europium Temperature Dependent Fluorescence Intensity Profiles
Evaluated with a Digital Camera and Spectrometer

37

The three tests showed good agreement with each other even though two independent
researchers used two evaluation methods over a three-year time period and used
fluorescent chemicals from different processed batches. Figure 2.8 shows that the
temperature dependent fluorescence characteristic of the Europium sample has a long
shelf life of at least 3 years and that the calibration profiles are in good agreement
independent of the experimental methods used to acquire the temperature dependent

fluorescence emission intensities.

Another test was performed on a new test specimen in order to authenticate the
validity of the calibration curves. The test specimen was a 3.5mm diameter cartridge
heater that was coated with the Europium compound and instrumented with a foil
thermocouple. The cartridge heater was cycled fiom 24°C to 80°C and gray scale
intensity images were taken at steady state incrementally as the temperature was
increased. The data were analyzed with the ratiometric methods described by Equation
2.1. A third order polynomial was fit to the calibration data with an R2 value of 1 in
Figure 2.8 and this was used to convert gray scale intensities measured on the cartridge
heater into temperature values. The comparison between the temperatures predicted by
the calibration curve fit (Figure 2.7) and those measured by the calibrated thermocouple
are presented in Figure 2.9. A single pixel without signal averaging near the location of
the thermocouple was used as the measurement point. A linear line was added to the
plot to emphasize the agreement between the predicted and measured values. This
experiment established confidence that the temperature dependent fluorescence behaved

in the same manner for multiple fluorescent samples.

38

 
 
     

Measured T (°C)
8 3 8 8 5‘ 8 8

N
O
__L_ __l_“....‘_._ _.._.._.,..-.L. _..L__ ,_i__. 4.-.]

.5
o
_ _L.__

 

o I __— "T'_'_ ‘—_T_"__—— Ta..._'_—’T—_— l I if T f "fl
0 10 20 30 40 50 60 7O 80 90
Predicted T (°C)

Figure 2.9: Comparison of FERIT Measured Temperature with Thermocouple
Temperature

2.6 Temperature Dependent Fluorescence of Terbium

As mentioned earlier, a literature search suggested that Terbium has temperature
dependent fluorescence intensity variation down to —170°C for Terbium dissolved in
solvents not used in this research. The Terbium fluorophore was prepared in a similar
fashion as the Europium compound and dissolved in an organic solvent (methyl ethyl
ketone) with PMMA pellets. The details of the fluorophore preparation can be found in
Appendix A. An aluminum substrate was coated with the Terbium-polymer film using
an airbrush and evaluated at temperatures below zero to establish the temperature
dependent fluorescence intensity correlation. For the particular application of interest,
temperature measurement down to —80°C was desired. The coated aluminum substrate

was mounted to a 5cm x 10cm hollow copper block through which dry Nitrogen gas,

39

cooled by a liquid nitrogen bath, was circulated. The flow of the cooled nitrogen gas
was controlled using a flow valve. By adjusting the flow velocity of Nitrogen gas, the
temperature of the test specimen could be lowered to temperatures below —100°C. Due
to the low temperatures, the test environment needed to be purged of water vapor so that
condensation did not create optical aberrations. Evacuation was accomplished by
enclosing the fluorescent specimen, camera, and illumination source inside a plastic
bag. The dry nitrogen exiting the cooling block was used to purge water vapor fiom the
inside of the bag. Fluorescence intensity images were acquired at steady state
temperatures in the same manner as the Europium experiments. A calibrated
thermocouple was mounted to the aluminum substrate for reference and ratiometric
methods, as described for the Europium compound, were used to establish a
temperature versus fluorescence intensity emission curve. Isothermal conditions were
observed when the thermocouple readout did not change in a 608 period during data
acquisition. Figure 2.10 shows the temperature dependent intensity profile obtained
with a digital camera for the Terbium sample (n=3). The temperature was lowered to —

100°C but little intensity change was observed beyond -80°C.

40

 

Normalized Intensity

 

 

0.3
0.2
0.1 a
. . a . e a v
-80 -60 -4o -20 o 20
Temperature (°C)

Figure 2.10: Temperature Dependent Fluorescence Intensity Profile of Terbium

A test was also conducted with the Terbium compound to evaluate the
calibration data with a different sample coated with the Terbium polymer film. The
plastic pintle shown in Figure 2.3 was used as the test specimen. The pintle is a part of
an automobile pollution control valve through which air flows. The efficiency of the
valve is subject to environmental conditions such as humidity and temperature. The
goal of this work was to make "microscale" whole field temperature measurements
down to cryogenic temperatures without obstructing the flow path. The material
presented next represents a feasibility study performed evaluating the capability of
using the FERIT method down to —80°C.

A temperature control system was developed so that the temperature of the
pintle could be stabilized between room temperature and —80°C. The pintle was

contained in a glass tube and temperature controlled dry nitrogen gas flowed into the

41

inlet and the exit flow emptied into the enclosed test area to facilitate vapor evacuation.

A schematic of the details of the temperature control system is illustrated in Figure 2.11.

 

 

g Mixing Box

 

 

 

 

 

 

 

Liquid Nitrogen
v
K J
C} / -\° *4
( Glass Tube \

 

1.13351].
vI

T-Pintle TA

Figure 2.11: Schematic of the temperature control system used to create a
cryogenic flow environment. Not shown are the UV excitation source and digital
camera.

Nitrogen gas was first cooled to ~196°C in a liquid nitrogen bath and then flowed

through a portion of tube that contained a cartridge heater to warm the nitrogen gas to a

42

desired temperature. The gas then flowed through a handmade mixing box that
promoted uniform temperatures upon entry into the glass tube containing the pintle.
Temperatures in the tube were monitored by thermocouples upstream and downstream
of the pintle. A third thermocouple was embedded in the center of the pintle. This
thermocouple was used as the reference surface temperature measurement for the
isothermal conditions when measurements were performed. Temperature was adjusted
by controlling the flow rate of nitrogen gas and the voltage applied to the cartridge
heater. A comparison between the embedded thermocouple measurements and the low

temperature FERIT measurements are presented in Figure 2.12.

 

 

s?
2
3
E
a
E
0
.-
t:
r:
in
IL.

-80 a

-100 -

Thermocouple Tenperature (°C)

Figure 2.12: Comparison between thermocouple measurements and FERIT
measurements from 24°C to —80°C. The linear line was added to emphasize the
agreement between the data sets.

The measured F ERIT temperatures are within 3°C of the thermocouple data. The
difference in temperatures was attributed to the difficulty with maintaining an
isothermal environment. The thermocouples located upstream and downstream of the

pintle indicated temperature spikes related to instabilities in the flow of Nitrogen gas.

43

Handheld thermocouple readers used for the test had slow response rates, which
probably produced a lag in the displayed temperature with respect to the actual
temperature of the pintle. In addition to the slow response of thermocouple readers, the
reference thermocouple was embedded within the pintle, creating another time lag for
temperature measurement. Temperature control by conduction, as was done during the
calibration tests, proved to be more stable than convection in validating the low
temperature capabilities of FERIT.

Another part of this work was designed to assess the ability to detect the onset of
condensation for a sample exposed to air with water vapor present. An aluminum
sample coated with the Terbium polymer was mounted to a cooling block and cooled
from room temperature to —40°C. Temperature measurements using F ERIT and image
processing software were made along a line defined across the coated sample. F igurc
2.13 shows fluorescent intensity images at 7.7°C with no condensation and at —40°C
with visible condensation formation.

No (.‘(illdCllsation COIlLlCllSLlllOll

‘

\

V V

 

Figure 2.13: Condensation Detection with the Terbium Polymer Film

 

The temperatures measured by the F ERIT method along the lines in the previous
images are presented in Figure 2.14. The data show that, as temperature increases, the
amount of apparent “noise” in the temperature measurements also increases. The
quantification of the amount of noise in the fluorescence intensity images that would
indicate the presence of condensation has not been performed. The data are presented
as preliminary evidence that the current fluorescence measurement techniques may be
useful with respect to quantitative detection of the onset and development of

condensation and frost formation.

 

Intensity (arbitrary unite

 

Line Length (mm)

——18.7°C . “-0.3°C --—---40°C

Figure 2.14: Ability of FERIT to Detect the Onset of Condensation at Low
Temperatures

The material presented concerning the low temperature measurement capabilities of
FERIT is not applied to any other application in this thesis. The technique has been
developed and its feasibility has been established. The final chapter of this thesis

describes future applications of the method.

45

2.7 Evaluation of the LEI fluorescence chemicals

Rhodamine B and Rhodamine 110 are used in their manufactured form unlike
Terbium and Europium, which must be manufactured by chemical processes in the lab.
The Rhodamines are mixed in defined quantities in a solute and are used as a
temperature indicator of the medium surrounding a heat source submerged in a liquid.
Rhodamine B is the temperature dependent indicator whereas Rhodamine 110 is
virtually temperature independent and acts as a reference to account for optical
aberrations. These chemicals emit at two different wavelengths. By taking the ratio of
the intensities of these two wavelengths, temperature can be quantified. The details of
this approach are presented later in the chapter. Figure 2.15 shows a radiofiequency
heating device submerged in saline solution mixed with Rhodamine B and Rhodamine
110 without illumination (A) and with the excitation source turned on (B). The

excitation source for the Rhodamine compound is the 488nm emission line of an Argon

Ion laser.

 

Figure 2.15: Rhodamine B and Rhodamine 110 mixed in saline solution without
excitation illumination and with illumination by an Argon laser sheet (488nm)

46

 

The concentration of the Rhodamines is such that the optical clarity of the solution is

unaffected.

2.8 Evaluation of the Spectral Characteristics of Rhodamine B and 110

Evaluation of the spectral characteristics of the fluorescent materials in this
thesis has been performed with a spectrometer. Two models of spectrometers were
used for evaluation of the Rhodamine fluorophores. The first model is the F -4500
fluorescence spectrophotometer (Hitachi Instruments, San Jose, CA) and the second is
the previously mentioned 82000 spectrometer (Ocean Optics). The F-4500 model was
used for the initial evaluation of the Rhodamine chemicals to determine the excitation
frequency that produces the maximum emission response from the compounds. The
spectrometer has the capability to excite the test sample at all wavelengths of
electromagnetic radiation from UV to near IR with 1 to 10nm bandpass widths. The
excitation wavelength that produces the highest emission response can be determined in
this fashion by scanning the excitation wavelength through a range of frequencies and
measuring the resulting emission response of the chemical. Figure 2.16 shows the
emission response of Rhodamine B with varying excitation frequencies. The left peaks
are the excitation frequencies. The plot shows that the maximum emission intensity

occurs using an excitation wavelength near 490nm.

47

4000 a
3500 l

 

A f\\
1% 3000 ! ,' Emission
=3 ., \ Waxelengths
5’ 2500 l Excitation .'
E ; Wavelengths . /
g 2000 a. \
g; 1500 l K// \\‘.
2 I \
a; 1000 l ‘a,
500 fij n ' \~\ ‘ z
o A . -__ . . - - i
450 500 550 600 650 700

Wavelength (nm)

—- 475 nm ——480 nm .-._ 490 nm —— 495 nm

Figure 2.16: Emission response of Rhodamine B at Various Excitation Frequencies

The same experiment was conducted with a pure Rhodamine 110 solution. Figure 2.17
shows that Rhodamine 110 is excited by wavelengths fiom 475nm to 495nm with
maximum emission occurring at 480nm excitation. Figure 2.16 and Figure 2.17 suggest
that a common excitation wavelength between 475nm and 495nm can be used to excite
both Rhodamine 110 and Rhodamine B. The most readily available coherent excitation

source is the 488nm wavelength of an Argon Ion laser.

48

1200 1

Emission
Wavelengths

       
 

1000-7
1

800 l

i

600 a

400 - Excitation
Wavelengths

Intensity (arbitrary units)

200 .-

Wavelength (nm)

—475nm ——480nm ——-490nm ———-495nm

Figure 2.17: Emission Response of Rhodamine 110 at Various Excitation
Wavelengths

2.9 LIFT Ratiometric Methods for the Determination of Temperature

The ratiometric techniques associated with the LIFT technique are similar to
those presented for the second ratiometric method in the FERIT section. The LIFT
method uses two fluorophores simultaneously in order to measure temperatures. The
two chemicals used are Rhodamine B, a temperature sensitive fluorophores, and
Rhodamine 110, a temperature insensitive fluorophores. The purpose of using two
fluorophores is to reduce error resulting from intensity changes caused by parameters

other than temperature. Since Rhodamine B and Rhodamine 110 fluoresce at different

49

wavelengths, any change in fluorescence intensity resulting from a non-temperature
related parameter will be seen in both emission intensities of the fluorophores. The
ratio of the emission intensities of Rhodamine B and Rhodamine 110 are taken as a
measure of temperature so two images are taken for one temperature measurement. An
automatic filter wheel under computer control was used to rotate a filter in place
corresponding to the emission wavelengths of Rhodamine B and Rhodamine 110.
Filters centered at 520nm and 580nm, lOnm bandpass, (Edmund Optics) were used in
the filter wheel. The computer coordinated the position of the filters with the
acquisition of the camera. At the start of an experiment, isothermal reference images
are taken at room temperature and used as the divisor for all other intensity ratios.
Equation 3 shows the operation performed to obtain temperature. The final 1mm value is

converted into temperature using calibration data.

 

IRHB . . I . . I . .

(U) RHBOJ) EXPO.»

"—_I 2’ I REFU, j) —1 3 I EXPO, j) I 3 1 ratio
RH110(i, j) REF RH110(l', j) EXP Ram,»

(23)

This method was suitable for steady state situations such as calibration procedures but
proved to be too slow when transient data were desired due to the lag caused by using
the filter wheel. For this reason, during transient acquisition only the Rhodamine B
intensity emission was recorded and a ratio exactly like Equation 2.2 was used. This
proved to be an effective means for fast temperature measurements. In order for the
ratio of the two emission intensities to be recorded in an effective manner, two cameras

are needed to operate simultaneously.

50

2.10 Temperature Dependence of Rhodamine B, Rhodamine 110 Solution

Evaluation of the temperature dependence of the Rhodamine B Rhodamine 110
solution was performed with the Ocean Optics Spectrometer. As stated earlier, the
spectrometer sample holder was connected to a constant temperature circulation bath
(N eslab). The bath can vary the temperature of the specimen in the sample holder from
—12°C to 100°C. The manufacturer of the spectrometer does not recommend
temperatures higher than 100°C. A solution of saline plus 0.05rngl'l Rhodamine B and
0.0025mgl'l Rhodamine 110 was placed in a quartz cuvette in the sample holder. The
Rhodamine B-110 solution was known to have strong temperature dependent
fluorescence from 15°C to 40°C.“;43 The Ocean Optics spectrometer was used to
establish the feasibility of using the Rhodamine B-l 10 solution to measure temperatures
up to 90°C for the measurement of temperature fields surrounding radiofiequency
heating devices. Previous results had only gone to 40°C. Starting from room
temperature, the temperature of the Rhodamine solution was increased by ten-degree
increments to 90°C. A thermocouple was mounted inside the quartz cuvette and a small
magnetic stir bar and stirrer was used to create an isothermal solution. A 488nm light
source was fabricated to excite the solution. The 488nm wavelength was chosen based
on the initial emission response tests of the chemicals and the anticipation that a 488nm
Argon laser will be used for the actual experiments. The next section describes the
details of the fabricated light source for these experiments. Figure 2.18 shows the
temperature dependent fluorescence of the Rhodamine B and Rhodamine 110 solution.
The spectral output of Rhodamine B is located at approximately 580nm whereas the

spectral output of Rhodamine 110 is located at approximately 520nm. Rhodamine B

51

exhibits fluorescence intensity decay as temperature is increased and Rhodamine 110
exhibits very little change with temperature variation. The intensity of Rhodamine B
decreased by 70% fi'om 20°C to 90°C and Rhodamine 110 decreased 3% from 20°C to
90°C. The Rhodamine 110 emission acts as an internal reference point to account for

fluctuations in fluorescence not related to temperature.

Intensity (Illmax)

 

Wavelength (nm)

——19.2°C —30.3°C * 38.4°C 7—51.8°C 63°C —74.3°C —90.1°C

Figure 2.18: Intensity variation of Rhodamine B, Rhodamine 110 solution

If there is a change in the test environment such as a fluctuation in the excitation
radiation source that affects the fluorescence emission of Rhodamine B then Rhodamine
110 should also experience a proportional change. By taking a ratio between the peak
emission spectra of these chemicals, any variation in intensity of the light source or any

other fluctuation should be removed.

52

The ratio of the intensity values of the Rhodamine B and 110 peaks are plotted against
temperature in Figure 2.19. Regression analysis is applied to these data with Curve Fit
2D® or Microsoft Excel® in order to convert experimental ratiometric intensity data into

temperature.

Intensity Ratio/lmax
O
at
I

 

24 34 44 54 64 74 84 94
Temperature (°C)

Figure 2.19: Fluorescence Ratio of Rhodamine B and Rhodamine 110 with Respect
to Temperature

When the actual experiments are performed, a digital camera captures gray scale
intensity images that are filtered to match the wavelengths of Rhodamine B and
Rhodamine 110. Two images are actually taken, each filtered with a 10nm bandpass
filter centered at 520nm and 580nm, to measure the intensities corresponding to
Rhodamine B and Rhodamine 110. The ratio of the gray scale values is converted to

temperature with a calibration similar to the one shown in Figure 2.19. Periodic

53

calibration was performed after a new mixture of Rhodamine B and Rhodamine 110
was made to ensure that the linear regression was still valid for a new fluorescent
solution.

It should be noted that the locations of the peaks in Figure 2.18 are different
from those of Figure 2.16 and Figure 2.17 for the excitation frequency evaluation of
Rhodamine B and Rhodamine 110. The concentration of the Rhodamine B and 110
was much higher than what was used for the results in Figure 2.18. The exact
concentration was unknown but a good measure to the extent of the concentration was
the color of the solution. The color of the first test specimens was red for the
Rhodamine B and yellow for the Rhodamine 110 whereas the solution used for
calibration in Figure 2.18 was clear to a faint pink color. The fluorescent solution
should contain the smallest amount of fluorescent chemical so that the chemical does
not act as its own quencher. Basically, the darker the solution the harder it is for the
emitted radiation to leave the solution in its original form. The fluorescent chemical in
solution reabsorbs the emitted radiation and diminishes the radiated energy resulting in
longer wavelengths exiting the test volume. A following section presents a discussion

on the effect of concentration on temperature dependence.

2.11 Fabrication of 488nm Light Source for Calibration of Rhodamine B-l 10
Solution

A special light source was designed to excite the Rhodamine B and Rhodamine
110 solution since the light source supplied with the spectrometer was not powerful

enough to produce a measurable emission response. This light source was used only

54

for the calibration experiments with the spectrometer. The light source consisted of a
lamp house (Dolan-Jenner) with a bulb that emitted electromagnetic radiation from
approximately 400nm to 750nm. A 488nm, 10am bandpass, (Edmund Optics) filter
was mounted in the light path to segment the desired wavelength of light. Figure 2.20
shows the spectral output of the light source without filtering and with the 488nm
bandpass filter in place. The plots are normalized with respect to themselves. The
intersection of the plots represents the portion of the unfiltered radiation that is being
used as the excitation wavelength. This light source can be used for a wide range of

excitation wavelengths using the appropriate optical barrier filters.

1 1
0.9i

Intensity/lg.“
.0 .° .0
03 :1 9°

9.0.0
(999*

.09
dN

 

AA

”A _..‘.A

200 300 400 500 600 700 800
Wavelength (nm)

:3
l

Unfiltered —-— 488nm bandpass filter

Figure 2.20: Evaluation of the Light Source Used for Calibration of Rhodamine B
and Rhodamine 110

The light was conveyed to the fluorescent test specimen in the spectrometer sample

holder with a 3mm diameter fiber optic cable instead of the 600m diameter fiber optic

55

that was included with the spectrometer. The light collimator was removed from the
excitation port of the sample holder so the larger fiber optic could be inserted into the
holder. The emission port was left unchanged. This light source made it possible to
analyze very low concentrations of Rhodamine B and 110 without using a high intensity

source such as an Argon laser, which will be used in the application experiments.

2.12 Effect of Concentration of Rhodamine B-110 Solution on Spectral Emission

It is important to understand the effect the concentrations of Rhodamine B and
110 have on the temperature dependence. It has already been shown that a spectral shift
exists for high concentrations of Rhodamine B in a solution from a comparison between
Figure 2.16 and Figure 2.19. A fluorescent chemical absorbs the radiation fi'om the
excitation source and emits the energy at a longer wavelength than that of the excitation
source. The emitted radiation is absorbed a second time by the excess fluorescent
chemical in solution creating another spectral shift before final emission fiom the test

3' Figure 2.21 plots the wavelength of the maximum intensity or peak of

volume.
Rhodamine B measured by the spectrometer due to variations in the concentration. This
plot shows an emission shift to higher wavelengths indicating an increased energy loss
with increasing concentration of Rhodamine B. The spectral shift observed in Figure
2.21 also plays a critical role in optical filter selection. The spectrometer measures the
intensity of all wavelengths but a digital camera must rely on filters to separate the
individual wavelengths of light. If the maximum peak wavelength is shifted to a longer
wavelength than the published wavelength then this is a good indication that the

concentration of fluorescent chemical is too high. The temperature dependence of these

56

concentrations was also tested to understand the impact variations in concentration

might have on a calibration curve.

588 t
.5 586.65 X
a
.2 l
E |
584 1
"i l A 583.41
g a
"6 E .
5 580 1
2’ I 578.87 1
.2
g l
0.01 0.1 1 10
Concentration (log mgll)

e 0.015mgll I0.045mg/| A 3.0mgll ><100mgn

Figure 2.21: Spectral Emission Shift Due to Concentration of Rhodamine B

The lower concentrations are more susceptible to variation of the fluorescent chemical
in the preparation of the solution since very small amounts of solute must be mixed with
the solvent. Larger quantities of the fluorescent solution could have been made to
reduce the frequency a new batch was made but other factors such as shelf life and
potential contamination by other means may have been possible. The temperature
dependence of Rhodamine B and 110 at varying concentrations is plotted in Figure

2.22.

57

1 $
0.9 - e
0.8 ~ x
0.7 1 ‘
0.6 1 e
0.5 . * .
0.4 ~ *
0.3 7 l A.
0.2 ~ ‘5 ,t >9
0.1 ~
0 l l l

25 45 65 85
Temperature (°C)
9 0.1rng/L (20 to 1) x 3.0rng/L (6 to 1) A 6.0mg/L (6 to 1)

Intensity Ratio/l...“

 

 

Figure 2.22: Intensity Ratio Dependence on Temperature with Varying
Concentrations of Rhodamine B and 110

The concentrations are denoted in terms of Rhodamine B and its proportion with
Rhodamine 110. All data were normalized with respect to the ratio of the intensity of
Rhodamine B to 110 at 25°C. Figure 2.22 shows that the normalized temperature
dependent intensity profiles are not affected by the concentration of the Rhodamine B
and Rhodamine 110, thus simplifying the LIFT method greatly. This means that any
quantity of Rhodamine B and Rhodamine 110 can be mixed in a fluid and temperature

measurements can begin immediately.

58

Chapter 3: Application of LIFT and FERIT to Measure Temperature
Fields On and Around Radiofrequency Heating Devices

3.1 Application of LIFT to Measure Temperature Fields Surrounding RF Probes

The measurement of the thermal fields surrounding radiofrequency heating
probes is accomplished utilizing the temperature dependent fluorescence of Rhodamine
B and Rhodamine 110 described in the previous chapter. Rhodamine B and Rhodamine
110 are excited by a large bandwidth of visible radiation wavelengths of light. In the
previous chapter, the emission response for the two fluorophores was shown for several
excitation wavelengths. This established a common excitation fiequency between
Rhodamine B and 110. The frequency of 488nm was selected since it produced high
emission responses from both the fluorophores and it is a common wavelength
produced by an Argon ion laser. The laser provides coherent, high-energy, narrow
bandpass, electromagnetic radiation that enables large volumes of fluid embedded with
fluorophores to be excited resulting in strong emission intensities measurable by
sensitive cameras.

A validation test was performed with an immersion heater, a non-radiofrequency
heating device, so that thermocouples could be used to verify the temperatures
measured with the LIFT technique. This was done before temperature measurements

on the actual radiofiequency probes were attempted to ensure that the LIFT technique

59

produced accurate results. Figure 3.1 illustrates the equipment used to validate LIFT.
A cylindrical lens, 3.5mm diameter, was placed in the path of an argon ion laser beam,
0.5W, to produce a laser sheet approximately 1 mm thick. The laser sheet passed
through the Rhodamine B and Rhodamine 110 solution and impinged on the centerline
axis of the submerged immersion heater. A similar setup was also used for the

radiofrequency probes.

Immersion Heater or
RF Probe

Argon laser ,. \T
488nm T’Cs

-"Ihser ($0

 

 

 

 

 

 

_ Rhodafitlne B,
Cylmdnml Rhodamine
Lens ...... 110 solution

 

 

 

 

Filter Changer [:1
Zoom Lens [:l

 

Law Ught
Intensity Digitall
Camera

 

 

 

Figure 3.1: Schematic of Validation Test

Four thermocouples, 0.255mm diameter T-type, were positioned as shown
approximately 4 mm from each other with the closest thermocouple being 0.5mm from
the immersion heater. A low light digital camera (Cooke SensiCam) measured the

emission intensity while a filter changer separated the emission spectra of Rhodamine B

60

and Rhodamine 110. From the previous chapter, Rhodamine B fluoresces at 576nm and
Rhodamine 110 fluoresces at 520nm, so 580nm, 10am bandpass and a 520nm, 10am
bandpass filters respectively were used in the filter changer. Reference images for both
emission spectra were taken at room temperature before the immersion heater was
turned on. After the immersion heater was powered, the thermocouples acquired time—
temperature data at le and the digital camera acquired fluorescence intensity images
at regular timed intervals, as the immersion heater produced temperature gradients in
the saline medium. The images were later matched to the time-temperature data
obtained by the thermocouples. During image acquisition, the filter changer
automatically changed the filters in between shots so that the ratio of the Rhodamine B
and Rhodamine 110 intensities could be performed after the test. Image analysis
software (Image Pro Plus®) was used to extract intensity values from each image at
every pixel along a line directly above the location of the thermocouples. Intensity
values were extracted fi'om corresponding Rhodamine B and Rhodamine 110 images
and spreadsheet software (Excel) was used to produce the intensity ratios for each pixel
along the line profile. A curve fit equation generated from calibration data was used to
convert the intensity ratios into temperatures. Figure 3.2 shows the comparison
between LIFT measured temperatures along a line profile with temperatures measured

at a point by thermocouples at corresponding times.

61

90
88 l ,0

on
0)

co on
N h

L4 ‘_4.*1A.k4, #1.
e

N
on

 

Temperature (°C)
00
O

\lxr
A0)

‘0' e e 0 e
v at» ”"A’.‘Ir‘~"“4~ ° ’.° M” u
" 3"“ 4 ‘M~.o‘yfv TC1
T93 Tcz “1‘83.

0

‘1
N

\l

0
1mm;
1
4
l
l

"Vi liW‘F fi’ili'fi V'V'7'#'fi fi'ifi

4 5 6 7 8 9101112131415
Distance from immersion Heater (mm)

0
_|
”1
(at)

Figure 3.2: Comparison of thermocouple measurements with LIFT temperature
measurements. Inset picture is a gray scale LIFT image showing the location of
the thermocouples and approximate location of the line profile.

The inset picture is an unprocessed intensity image of the emission of Rhodamine B.
The immersion heater is on the right side of the picture and the thermocouples are
extending to the left. The intensity of each pixel along the line profile was divided by
the corresponding pixel fiom the corresponding Rhodamine 110 intensity image. The
location of the thermocouples was spatially resolved with the location of the line profile
for direct comparison of the two temperature measurement methods. The maximum
acquisition rate obtainable by the imaging equipment for this validation experiment was
limited to approximately 0.5 Hz. The filter changer and the relatively long exposure
times of 0.755 for each image limited the acquisition rate. This did not prove to be a
problem for this experiment since the temperature of the fluid was changing at a rate

close to 0.027°C/s. The data presented in Figure 3.2 show agreement between LIFT

62

and thermocouple temperature measurements to be, on average, within O.88°C of each
other. The maximum difference between thermocouples and LIFT occurred at the
location closest to the immersion heater, which was the location of the steepest
temperature gradient. At this point, the maximum difference between the two
measurements was observed to be 1.5°C. Pixel averaging reduced the error in the
region closest to the immersion and did not have much of an effect on the region farther
away fi'om the heater. This was probably due to the small gradients that existed away
from the immersion heater. This test established confidence that LIFT can be used as
an independent means for future measurements around radiofiequency probes when

there is no other means to validate the LIFT measurements.

3.2 Image Acquisition Rate Improvement

The previous section indicated an imaging acquisition rate of 0.5 Hz. It is the
goal of this thesis to obtain image acquisition rates close to 10 Hz. There are several
things that can be done to improve acquisition rates. The first includes increasing the
concentration of the fluorophores, which in turn increases the fluorescent emission
intensity. This is effective up to a certain point. As the concentration of the
fluorophores is increased, the optical clarity of the solution decreases making it more
difficult to see a submerged heat source. A concentration of Rhodamine B between
0.05mg/l and 0.15mg/L is recommended. The ratio of concentrate between the two
fluorophores was always kept at 20:1 so the amount of Rhodamine 110 recommended is
20 times less than that of Rhodamine B. This ratio was chosen based on work done by

other researchers.42 Figure 3 .3 shows the effect concentration of the fluorophores has

63

on the measured emission intensity at a constant integration time. Integration times
were chosen so that the measured intensity output of the lowest concentration was near
1000 counts out of 4096. A solution of 0.05mg/l Rhodamine B and 0.0025mg/l
Rhodamine 110 was used as the starting point. The concentration of the solution was

doubled and tripled and evaluated at the same integration times with the Ocean Optics

spectrometer.

3000 7

’ 2647.905
2500 I

I

I
2000 e

1 738.143

1500 7 692.524

475.286
1000 - 252.333

Intensity (arbitrary unlts)

   
 

 

 

 

 

 

 

500 < ~
0 — _ ——r~— e I A - I
500 550 600 650 700
Wavelength (nm)
1x ~ 2x 3x

Figure 3.3: Effect of Concentration of Rhodamine B and Rhodamine 110 on
Emission Intensity

Figure 3.3 indicates emission intensity values for the peak at 520nm (Rhodamine 110)
and the peak at 576nm (Rhodamine B). These values essential increased in direct

proportion to the fluorophore concentration. This suggests that the integration time can
be reduced three fold by increasing the concentration of the solution from the 1x

mixture to the 3x mixture. Since very small quantities of fluorescent chemical are being

64

mixed with relatively small volumes of fluid, an exact doubling or tripling was not
expected with doubling or tripling the concentrate due to uncertainties in the
measurement of the small quantities of fluorophore.

The intensity of the laser or power of the incandescent source used to excite the
fluorescent solution can be increased in order to produce higher emission response from
the fluorophores, thereby reducing the integration time needed by the detector to
measure a signal. The downside to higher excitation intensities is that the fluorophores
tend to quench at faster rates resulting in degradation of fluorescence intensity not
resulting from a temperature change. Adrian42 indicates that exposing a 3m] cuvette to a
laser power of 0.5 W for duration of 1208 results in negligible decay of fluorescence.
The volume of fluid used for the experiments in this thesis was approximately 500ml,
so it was concluded that the quenching effect of the laser at the same power setting was
less than Adrian observed. Lasing of the fluorescent solution was limited to 1203 and
the power was never increased beyond 0.5W.

A third parameter that resulted in improved acquisition rate was the bin setting
of the digital camera. The 0.5 Hz rate images were taken with no binning performed to
the image. Binning reduces the resolution and amplifies the measured intensity. For
example, a 2x2 bin will add the four nearest neighbor pixels together and replace the
values of all four with the value of the sum. An image that was originally 1280 x 1024
pixels is reduced to an image of 640 x 514 pixels. The binning size used in this
research was an 8 x 8 bin. This reduced the image to 160 x 128 pixels and provided a
resolution of 90m or 9 pixels per mm. This provided adequate resolution for the

experiments.

65

The three parameters mentioned were controlled to provide the fastest integration
times for the digital camera while maintaining adequate image resolution and optical
clarity of the saline solution. The concentration of Rhodamine B was never increased
above 0.045 mg/L and the laser power was kept at or below 0.5W, while binning did
not exceed 8 x 8. The SensiCam digital camera was able to capture images at
approximately 10 Hz mainly due to the 8 x 8 bin. This parameter had the most effect on
integration speed and was the easiest parameter to adjust. The 8 x 8 binning made it
possible to image with lOms integration times. This would have produced a 100 Hz
framing rate but this limit was reduced by the data transfer speed from the camera to the

computer.

3.3 Field Temperature Measurements Around RF Probes

The LIFT method has been applied to quantify the temperature fields surrounding
RF probes submerged in saline solution. The probes were held in the same fashion as
illustrated in Figure 3.1 for the validation experiment with the immersion heater.
Figure 3.4 shows an image of the radiofrequency probes and details of the probe tips.
The monopolar probe on the left has a smooth tip surface and a 2.3mm diameter,
whereas the bipolar probe on the right has a 3.0mm diameter and many holes evenly
spaced in the tip. The reason for the holes is unknown. Initial tests with the LIFT
method were conducted at flaming rates near 1 Hz since the improvements in

acquisition rate mentioned in the previous section were not yet implemented.

    

Monopolar Bipolar

Figure 3.4: Image of monopolar and bipolar radiofrequency probes. Close up
images show the details of the probe tips.

For these experiments, the probe was operated in an optically clear tissue phantom gel.
The characteristics of this gel are discussed below. A gel or solid medium was desired
so that heat transfer would occur by conduction only, which would make it easier for
eventual calculation of the heat flux from the probe tip. Another reason was to
eliminate potential optical aberrations caused by convective flow currents and bubble
formation resulting from the fluid medium de-gassing at high temperatures. De-gassing
is the phenomenon witnessed before a fluid is brought to boil. As a fluid such as water
is heated up, bubbles form on the surface of the container holding the fluid well before
boiling occurs. This phenomenon was witnessed during the preliminary operations of
the RF probes. In anticipation that the bubbles will obstruct the fluorescence emission
intensity measurements, a gel phantom was created. The desired properties of the gel
phantom are such that it is optically clear so that temperature can be measured within

the medium. The gel must also be a good electrical conductor so that the RF probes

67

operate in the medium the same way that they operate in pure saline solution.
Rhodamine B and Rhodamine 110 must be soluble in the medium. The last desired
parameter is that the gel phantom has a high melting temperature greater than 90°C.
However, this goal could not be realized at this time. The gel phantom was created by
dissolving 30g/L gelatin (Sigma-Aldrich) in addition to the Rhodamine B and
Rhodamine 110 fluorophores in saline solution. The new solution was a solid at
temperatures less than 40°C. When in liquid form it had a higher viscosity than the
saline solution. The gel medium was also optically clear as denoted by Figure 3.5. The
probe can easily be seen as well as the camera lens on the other side of the glass test
container. An important finding was that the addition of gelatin to the saline solution

did not affect the calibration curves established in the previous chapter.

 

 

Figure 3.5: Image of monopolar RF Probe Submerged in an Optically Clear Gel
Medium

Figure 3.6 is an image of the test equipment used for the LIFT experiments. The
path of the Argon ion laser (Lexel Laser model 95) is denoted by the arrows. The laser

beam is made into a sheet before impinging on the fluorescent medium containing

68

Rhodamine B and Rhodamine 110. The cylindrical lens, which transforms the laser

beam into a sheet, is not observable in the image.

    
   

“s-«ewwnsewmm v, _,t..,..,..,._,..'... w

72:; area. 16
run '

Argon Ion Laser g5

   

Figure 3.6: Experimental Equipment for LIFT measurements indicating the laser
beam path, RF probe submerged in an optically clear tissue phantom, RF
generators, and digital camera.

The bipolar RF generator, ArthroCare Corp., is on top and the monopolar RF generator,
Oratec Interventions, is located on the bottom. The digital camera, Spectra Source
Orbis, is located at the rear of the image. This camera was used in the preliminary
studies but was switched to a newer camera with higher quantum efficiency, Cooke
SensiCam, for experiments later in the research. The new camera made transient

temperature analysis possible.

69

3.4 Results of LIFT measurement in gel medium

3.4.1 Monopolar

LIFT images were taken at various times during a 90 second period while the
probe was powered at a control unit setting of 30 W and a probe tip temperature setting
of 75°C. Figure 3.7 shows temperature variation along horizontal (x-direction) and
vertical (y—direction) directions. The locations of the lines along which temperature
profiles are measured are shown in the inset image. Note that the ground plate for the
monopolar RF probe was mounted on the bottom of the test section directly below the

tip of the probe as was illustrated in Figure 1.1.

 

     

x-direetion

00000

Temperature (°C)
8 3 it. 8 it

 

 

 

30
0 5 10

Distance from RF probe tip (mm)

Figure 3.7: LIFT Predicted Temperature Around the Monopolar RF Probe

Before the experiments, it was not known whether the LIFT technique would produce
reliable data near the probe tip due to the presence of small convection currents and the

reflection of the laser light from the probe surface (both of these factors were believed

7O

to have the potential to distort the LIFT measurements). It was observed during the
experiments that the addition of gelatin to the solution served to reduce but not
eliminate the convection currents. The laser light reflected from the probe surface also
did not appear to be a major factor. For the monopolar RF probe, there were no
observable optical aberrations in the image close to the probe tip. As seen in Figure 3.7,
with a monopolar RF probe, 8 more pronounced thermal effect was observed in the
vertical (y) direction than in the horizontal (x) direction. It is therefore obvious that the
shapes of the isotherms around the probe are not symmetrical (for distances greater than

1mm).

3.4.2 Bipolar

Similar to the experiments performed with the Monopolar RF probe, LIFT
images were taken at various times during a 90 second period while the probe was
powered at a control unit setting of “1”. The x-direction and y-direction line profile
results from the bipolar RF probe are presented in Figure 3.8. Contrary to the
monopolar probe examined, air bubbles formed very close to the tip of the bipolar RF
probe, creating optical disturbances that affected the LIFT temperature measurements.
The edge of the probe tip could not be distinguished clearly due to the aberrations. The
size of the bubbles was less than 0.5 min. Due to the presence of the bubbles, the data
presented in Figure 3.8 are displaced away fiom the surface of the probe tip, beyond the
location of the bubbles. When Figure 3.7 and Figure 3.8 are compared, it is observed
that the temperature rise (in any direction) with the monopolar RF probe is greater than

the bipolar. This observation is in very good agreement with published in vitro

71

experimental studies and numerical simulations 44. The isotherms around the bipolar RF

probe tip are more symmetrical than the monopolar case.

 

 

Temperature (°C)

 

 

 

 

Distance from RF probe tip (mm)

Figure 3.8: LIFT Predicted Temperature Around the Bipolar RF Probe

It was evident from the data that the bipolar probe created higher temperatures
in the x-direction, whereas the monopolar probe created higher temperatures in the y-
direction. This could be due to the location of the return electrode. For the monopolar
RF case, the return electrode is placed under the tissue being heated. Therefore, the
radiofrequency current flow from the probe tip through the tissue to the retum electrode
creates a volumetric heat generation within the target tissue. Due to these factors, it is
reasonable to assume that the temperatures in the vertical direction will be higher,
which is in agreement with the results presented here. Future experiments will be
performed to test this hypothesis and to determine the effect that the location of the
return electrode has on the temperature distribution in the surrounding medium.

The return electrode for the bipolar RF probe, as was illustrated in Figure 1.1, is

located above the probe tip. Therefore, heat is generated in the saline or gelatin media

72

surrounding the probe tip and then is convected or conducted to the target tissue. It is
essentially a surface heat transfer phenomenon. It is thus reasonable that the maximum
temperature reached would be in the horizontal direction.

After these tests were conducted, an experimental design was established that
more closely matched the actual operating conditions of the RF probes. Instead of
using an entirely gel medium, a layer of gel was covered with saline solution. In this
fashion, the gelatin layer represented a tissue phantom in which the probe was brought
into light contact with the gelatin surface. Enough saline solution was added to
submerge the probe tips to ensure proper operation of the RF probes. Figure 3.9
illustrates the configuration where the gel phantom is on the bottom with saline solution

covering the phantom and submerging both RF probes in the fluid medium.

Bipolar Monopolar

   

Saline

 

fi Gel Phantom

 

 

 

 

'——* Ground Plate

 

 

Argon Laser
488 nm

 

 

Figure 3.9: Gel phantom experimental configuration with the laser sheet passing
through the bottom of the glass test container.

73

Both probes were not operated at the same time but are shown in the diagram to
illustrate the relative distance the probes were operated from the gel surface. The saline
solution and gel medium both contained Rhodamine B and Rhodamine 110 so that
temperature in the both mediums could be measured. The laser sheet was brought
through the bottom of the glass test container in order to illuminate both sides of the
probe so that complete two—dimensional isotherms can be measured around the probe
tip. The first setup brought the laser sheet in fi'om the side so that the probe cast a
shadow on one side of the probe, rendering LIFT useless in the shadowed region. The
RF monopolar probe on the right is operated while in contact with the tissue phantom.
For this experiment, the ground plate was mounted on the side of the test container
instead of the bottom so that the laser sheet would be unobstructed. The bipolar RF
probe on the left is operated l to 2mm from the surface of the tissue as described by
surgeons who use the RF probes. The settings for both probes were kept the same as
those used with the first experiments. The data acquisition speed was improved to
10Hz with the adjustment of the parameters mentioned beforehand. Instead of using the
ratio of the intensity of Rhodamine B to the intensity of Rhodamine 110, only the
intensity of Rhodamine B was used so that fast acquisition could be accomplished. If
both images were taken, the flaming rate would be reduced due to the time needed to
switch the barrier filters in between image acquisitions. This single wavelength LIFT
method was presented in Chapter 2 and proved accurate for relatively short time
periods. The intensity of the laser proved to be very stable so intensity fluctuations due
to variations in laser power were not evident but other optical aberrations do have the

potential to disrupt the experiment.

74

Before testing began on the RF probes, the fluid and gel mediums were allowed
to stabilize at room temperature and a thermocouple was used to measure the
temperature of the fluid and gel medium. This temperature was used as the reference
point when converting measured emission intensity into temperature. A reference
Rhodamine B image was taken before the probe was turned on. Image acquisition was
performed at 10 Hz for a time period of 1205. Matlab code was written to read each
image and apply a regression analysis to each pixel in the image to covert measured
emission intensities into temperatures. The Matlab code also produced movies in *.avi
format. The code is presented in Appendix C.

The first probe evaluated was the monopolar probe. Figure 3.10 presents
individual frames acquired during the test for a probe set point of 30W, 75°C. The
color scale on the right side of the images ranges from 20°C to 80°C. Image A was
acquired l to 2 seconds after the probe was activated. Note the vortices rising along the
probe sides. These convective currents were characteristic of the monopolar
measurements shortly after activation. Image B was acquired after 45s of probe
activation. Temperatures in the gel reached nearly 60°C and all temperatures higher
than 40°C were assumed to be regions of liquid gelatin. The monopolar RF probe
produced symmetric isotherms with more thermal penetration in the vertical direction
than in the horizontal direction. These results are in agreement with the results

presented earlier in Figure 3.7.

75

 

Figure 3.10: Whole field temperature measurement in a gel and saline medium due
to monopolar RF heating. Line indicates gel-saline interface. Image A is l to 2
seconds post activation and image B is 45 seconds post activation.

A direct comparison between the two tests was not made since the operating mediums
were different for the two tests. The monopolar probe has a T-type thermocouple

embedded in the stainless steel tip, which is used as a reference for power control of the

76

probe. The thermocouple signal is filtered to remove the interference caused by
radiofrequency energy. It is assumed that differences in the thermal properties of the
operating medium, or the size and mass of the target object affect the power control of
the probe. This effect was tested by operating the RF monopolar probe in saline
solution only and comparing it to the results just presented in Figure 3.10. A RF
generator for the monopolar probe was acquired that had RS 232 outputs for computer
acquisition of power and temperature performance of the probe. Figure 3.1] shows the
reported power for the situation where the probe was operated in a gel-saline

environment and for the situation when the probe was operated in saline solution only.

Power (W)

 

 

 

 

 

0 — » , — ~ .,
20 21 22 23 24
Time (s)

 

 

—gel —— saline orly

Figure 3.11: Measured power output for the RF monopolar probe in a gel-saline
medium and in a saline solution only medium.

The results in Figure 3.11 indicate that the saline solution environment required more

“on time” than that of the gel-saline medium. A probable cause for this is that the probe

77

needs to stay powered longer for the saline solution medium since heat is transported

away from the probe tip more than when the gel medium is present.

The same procedure was followed for temperature measurements around the
bipolar radiofiequency probe at a set point of ‘1’. The results are presented in Figure
3.12 for 1 to 2 seconds post activation (image A) and for 100s post activation (image
B). The line represents the gelatin-saline solution interface, which was approximately
2mm below the bipolar probe tip. The arrows in image B indicate bubble formation.
The bipolar probe was plagued with the formation of bubbles much more than the
monopolar probe tip. The two probes were operated at comparable settings so one

probe was not experiencing much higher temperatures than the other one.

78

 

Figure 3.12: Whole field temperature measurement in a gel and saline medium due
to bipolar RF heating. Line indicates gel-saline interface. Image A is 1 to 2
seconds post activation and image B is 100 seconds post activation. Arrows in
image B indicate bubbles.

A possible reason for the bubble formation might be due to the contours of the probe
tip. The bipolar probe has many holes located on the probe tip and in general is not as

smooth as the monopolar probe. Rough surfaces and surfaces with small cavities are

79

reported to enhance heat transfer and heat flux.45

The porous surface facilitates vapor
formation and is ofien used for heat transfer enhancement during boiling. A gel
medium was used to attempt to remove the formation of bubbles but this was

unsuccessful.

3.4.3 Effect of Bubble Formation on LIFI‘ Measurements

A study was performed to quantify the effect a bubble has on the LIFT method.
Even though the bubbles were most prominent on the bipolar probe, the monopolar
probe was used as an example since a clear example of bubble formation to bubble
release was found. This test was performed in saline solution only at the 30W 75°C set
point. The images in Figure 3.13 show the bubble formation (A), bubble growth (B),
and bubble release (C). Four points around the probe were selected and time dependent
LIFT measurements were extracted from the images. One of the points was located at
the center of the bubble. This was done to examine the effect a bubble has on LIFT

measurements.

      

Bubble forms Bubble released

Figure 3.13: Bubble formation on RF monopolar probe tip. Bubbles create an
optical disturbance that effects LIFI‘ measurement.

8O

Figure 3.14 shows the time dependent LIFT measurements at the five locations
indicated on the inset picture. The closest points were within 1mm of the probe tip
whereas the furthest point was approximately 3mm from the probe tip. The data show
temperatures cycling at approximately 1 Hz between 30° and 40°C at the points closest
to the probe tip. The point furthest away from the probe fluctuated between 25°C and
30°C. The temperature data from the monopolar probe unit are also plotted and
represented as the point on the probe tip. The data measured by LIFT and that
measured by the RF control unit agreed with each other with respect to the fiequency of
temperature cycling. This indicates that the LIFT method can track 1 Hz temperature

fluctuations over a 10°C/s heating and cooling rate.

”7

70

Temperature (°C)

 

Figure 3.14: Temperature measurement effect of bubble formation. The points
around the probe tip represent the relative location of the measurements.

81

The data at the location of the bubble show an apparent temperature decrease as the
bubble is formed and a temperature measurement restoration once the bubble is
released. The formation of bubbles can potentially cause a temperature error in either
direction because the spherical geometry of the bubble can decrease the fluorescence
intensity resulting in a measured temperature increase or intensify the fluorescence
resulting in a measured temperature decrease. The LIFT measurement was still able to
track the temperature fluctuation rate through the bubble indicating that it may be
possible to correct the error by using the corresponding intensity signal from the
Rhodamine 110 fluorophore. The Rhodamine 110 intensity signal was not measured
for this experiment since fast acquisition rates were desired. It is assumed the bubble
would cause a corresponding decay or increase in fluorescent intensity signal for the
Rhodamine 110 fluorophore. Ifthis proves to be true, then the ratio of the intensities of
Rhodamine B and Rhodamine 110 can potentially eliminate optical aberrations caused
by the bubbles. The measurement of the intensities would need to be performed
"simultaneously". Two cameras optically focused on the same region of interest is one

method to attempt this correction.

3.5 FERIT Measurements on the Monopolar RF Probe

The F BRIT method was applied to measure temperatures on the surface of the
monopolar probe tip for a probe set point of 30W 75°C. This method was only applied
to the monopolar probe since the bubble formation on the tip of the monopolar probe
was significantly less than the bipolar probe. Future tests will include measurements on

the bipolar probe once an effective method has been established to handle the added

82

challenge of imaging through bubbles. The ratiometric method presented in Chapter 2
using a single reference image was applied. Prior to the actual test, the probe coated
with the Europium polymer was calibrated in the glass test chamber by incrementally
increasing the temperature of the fluid in the test vessel and taking images when
isothermal conditions were realized as judged by thermocouple measurements in the
fluid medium. The thermocouple in the probe tip was also utilized during the
calibration procedure for reference. Images were acquired at 10Hz while the probe was
powered. A Matlab program similar to that used in the LIFT measurements converted
fluorescence intensity into temperature (details in Appendix C). Figure 3.15 presents 2
sequential images in time showing the probe increasing in temperature during one of its

power cycles. The temperature scale for the images was 20°C to 90°C.

83

 

Figure 3.15: FERIT Measurements on the Surface the Monopolar Probe Tip

The results obtained with FERIT were compared to the data output from the RF
generator. This compMson was made by extracting temperature data averaged over a 3

x 3 pixel area at the center of the probe tip for each FERIT image and plotted with

respect to time in Figure 3.16. The data are in good agreement with each other with
respect to the cyclic power on and power off temperature control behavior of the RF
generator. The first eight seconds showed the FERIT measurements measuring
temperatures lower than the output temperature from the generator during the power off
portion of the cycle. This difference diminished as time increased and the maximum
and minimum temperatures measured by FERIT eventually matched those produced by

the generator or R8232 data.

Temperature (°C)

 

 

Tine (a)

— Oratec - RS 232 — Fen‘t

Figure 3.16: Comparison of FERIT Measurements with the Temperature
Measured by the Thermocouple Embedded in the Monopolar Probe Tip

The line drawn on the plot in Figure 3.16 represents the set point of 75°C entered on the
control panel of the RF generator. The data show that the generator maintained an
average of approximately 75°C by cycling the temperature between 60°C and 85°C.
The temperatures measured by LIFT in the saline solution at locations within 1mm of
the probe tip indicated oscillating temperatures ranging from 30°C to 40°C. This

suggests convection quickly removes the heat generated by the RF probes.

85

Chapter 4: Conclusions and Future Work

The majority of the work for this research was spent testing fluorescent
chemicals in order to understand their temperature dependent fluorescence
characteristics and using the techniques developed performing measurements on
radiofrequency heating devices and for applications where cryogenic temperatures are
present. The next step would be to use the temperature measurement techniques
developed in order to study the effect that the measured temperature has on its
surrounding environment whether it is a biological application or some other area where
conventional temperature measurement methods are not suitable. The FERIT and LIFT
methods are the only methods in existence capable of whole field temperature analysis
within a fluid medium or on a surface submerged in a fluid and for whole field analysis
at cryogenic temperatures. This benefit alone opens the door to perform detailed

experiments on everything related to cryobiology and to radiofrequency heating.

The results presented in the previous chapters showed the LIFT and F BRIT
method could be used to make temperature measurements at an acquisition rate of 10
Hz, with a maximum error of 1.5°C, and with 90 um spatial resolution. This spatial
resolution is limited by the optics used. The fundamental limit is the wavelength of
visible light, which is on the order of 0.5urn. The LIFT method, which was previously

developed by other researchers for measurements from 24°C to 40°C, has been

86

extended to measure temperatures as high as 90°C. The F BRIT method previously
developed under the supervision of the principal investigator has been extended fi'om a

range of 24°C to 80°C down to —80°C.

Accuracy can be improved by implementing a dual camera system so that both
emission spectra can be measured for the LIFT method. The FERIT method can benefit
fiom this too by possibly combining Europium with another fluorescent chemical that

may be temperature invariant.

Fluorescent thermography methods offer many benefits over conventional
temperature measurement techniques and at the same time are not appropriate to use for
certain situations. Implementing a fluorescence thermography system is not an easy
endeavor and the benefits of the method should definitely outweigh the time required to
learn the method. Fluorescence thermography has been compared to some other
conventional methods in Table 4.1. Thermocouples, fluoroptic thermometry, infrared
thermometry are all relatively easy methods to implement. Someone new to those
methods can make accurate measurements in less than 1 day. However, fluorescent
thermography methods require months for someone who has little knowledge of the
method. The most time consuming part is to establish confidence in the calibration so
that the user will have assurance the results are accurate when actual measurements are

made independent of other conventional methods.

Fluorescent thermography methods are limited by acquisition speed since the
fluorescent emission signal from the fluorophore is sometimes weak, requiring long
integration times by sensitive digital cameras. The digital cameras are also limited by

the data transfer rate to the computer. The performance of the digital cameras is

87

expected to improve with time so the acquisition rates will continue to improve with
technology advancement. In order achieve 10 Hz flaming rates, binning was performed
on the image, thereby reducing the spatial resolution. Acquisition speed will also

improve as cameras become more sensitive to low intensity light.

The limitations of the other temperature measurement methods are the situations
that make fluorescence thermography appealing. Fluorescence thermography is limited
by the equipment necessary for application since it requires several components. This
will make application of fluorescence thermography in a confined space difficult.

Another limitation of fluorescence thermography is the cost.

 

 

 

 

 

 

 

     
  

 
   
 

Level of Acquisition Accuracy: emp. patial imitations:
Difficulty:‘Speed: nge: esolution:
Thermocouple Low [High 1-2°C -240°C to Low IRF fields,
”00°C (NA) Corrosive
chemicals
Fluoroptic Low w (4 Hz) 1°C -200°C to Low ize of
Thermometry +450°C (NA) robes, Acq.
speed
Pfrared ‘Moderate rModerate to 1°C 40°C to Opaque to
hermography ' +2000°C uids, Low
nd temp

 
 

 

 

 

 

 

  

Table 4.1: Comparison of conventional temperature measurement methods with
fluorescent temperature measurements.

88

Current cost of the LIFT experimental equipment was $55,000. If dual camera
measurements are desired, another $15,000 should be added to the cost. This is cheaper

than some IR thermography units that can cost over $100,000.

4.1 Future Developments for Application of LIFI‘ and FERIT to RF Heating

The development of the LIFT technique makes it possible to explore the effect
many variables have on the outcome of thermal therapies applied to biological material.
In clinical applications, saline solution is circulated in the cavity where the target tissue
is being treated to avoid the formation of hotspots and to wash away debris. Future
tests include circulating the saline around the probe tip at different flow rates to
determine the best rate that prevents overheating of the tissue. In addition to flow rate,
probe sweep speed is a deciding factor that determines the amount of heat transferred
into the material being treated. Presently surgeons operate the probe in a manner that
they perceive to be visually effective. Quantifying the effects of this parameter will
enable surgeons to perform more effectively by defining probe velocities that may be
detrimental to the patient. A test configuration similar to the gel medium covered with
saline can be used in conjunction with a mechanical arm that sweeps the probe over the
tissue phantom surface. The depth of thermal penetration can then be quantified with

probe sweep speed as well as with the RF control unit settings.

Beyond the clinical applications, developers of RF probe units can verify if their control
parameters are producing the desired temperature fields. Currently RF probe

manufacturers lack the thermal detail measured in this thesis.

89

A recent attempt was made to perform LIFT and F ERIT simultaneously during
an experiment. This would make it possible to directly measure the heat flux from
probe tip into the heated medium. Problems arose flying to keep the emission spectra of
Rhodamine B and Europium separate. Nonetheless this possibility should be explored

in the future.

The combination of data provided by the LIFT and F ERIT methods should
enable accurate boundary conditions to be determined for finite element simulations
(FEM). Finite element models can even be set up to simulate the tissue phantom
experiments. If good agreement between experiments and FEM can be obtained, then
the material properties in the model could be changed to represent biological material
with the assumption that the FEM model is accurate based on the validation with the

tissue phantom experiments.

4.2 Future Application for the Low Temperature FERIT Method

A future application of the cryogenic F BRIT method is to apply the polymer
film to the glass window located in the dorsal skin flap chamber and monitor
temperatures on the surface while conducting cryosurgery. Figure 4.1 shows a sketch
of the dorsal skin flap chamber. The image to the right is an image of the cancer cells
inside the window chamber. The cancer cells have been genetically marked with green
fluorescent protein (GFP), tagging them as bright green spots in the inset image. Blood
vessels are also visible in the window chamber as dark channels. If detailed

temperature measurements were made on the glass window surface using the F ERIT

90

method and Terbium during cryosurgery, these measurements could be used to correlate
temperature to the type of tissue damage observed beneath the window chamber. The
cancer cells can be destroyed either by rapid freezing rates, which cause the cells to

burst, or by destroying the blood vessels nearby the cancer cell clusters, which will

 

starve the cells.
xx
<2? ‘
I. u
.1 l ‘.
WNC
e‘ t ,
l J. ,va ._;:._
#- J""‘
at. I...

Figure 4.1: Potential application of low temperature FERIT to measure
temperatures on the window chamber during cryosurgery in order to correlate
temperature and thermal damage.

The polymer coating should also be semitransparent so that the cells and blood vessels
can be observed with a microscope. Since IR cameras cannot measure temperatures
down to cryogenic levels and thermocouples are invasive, the FERIT method could lead
to great advancements for correlating temperature and cooling rates to cellular death

and vascular injury.

The intended use of GFP is as a marker for cell viability. If a cell dies so does
the fluorescence associated with the GFP. During cryosurgeries the fluorescence
intensity of the tumor was observed to increase as the freezing protocol was applied to

the tissue in the window chamber. Figure 4.2 illustrates the qualitative increase in

91

fluorescence intensity as the cancer cells are frozen, and a decrease in fluorescence as

the cells return to their normal temperature.

(ii-'1’ tumor before free/c

I’rozcn (il-‘P tumor

   

GH’ tumor duringl thzm

Figure 4.2: GFP Tumor Qualitative Intensity Measurements During Freeze-Thaw
Cycle

A pilot test was conducted to evaluate the temperature dependence of the
fluorescence intensity of GFP. Pure GFP was placed in a cuvette in the spectrometer
sample folder and analyzed from —12°C to 60°C. The 488nm light source used to
evaluate the Rhodamine B and Rhodamine 110 solutions was used to excite the GFP
solution. The temperature of the solution was decreased as low as the Neslab bath
could obtain. The results presented in Figure 4.3 show that GFP exhibits temperature
dependence from -12°C to 60°C. Further experimentation needs to be conducted to
determine the entire temperature dependent fluorescence range of GFP. GFP has the

potential to be an effective cellular temperature indicator.

92

1.2 ‘.

A . 1 i
5 .
E 0.8 4 0
a 1 ,
g 0.61 . .
g e
3 0.4 i,
E 1
0.2
i” ~— _H—“ ’ —o ‘i‘_ H —" A“ "W” _ TFMT— _ " #" b—‘I H"“_’_ m-..,_._. — —_ .._1
-20 0 20 40 60 80

Temperature (°C)

Figure 4.3: Temperature dependent fluorescence of GFP from -12°C to 60°C.

93

Appendix A: Procedure for Terbium and Europium Preparation

Chemical Preparation:

Terbium (III) Chloride hexahydrate and Europium (HI) Chloride hexahydrate must be
chemically altered before being used as a temperature dependent fluorophore. The
chemicals appear as a "salt" in their unprocessed form and do not exhibit any

fluorescent characteristics.

1. Mix 0.587 grams of Europium (Ill) Chloride or 0.598 grams of Terbium (III)
chloride in 50ml of distilled or deionized water.

2. Increase the pH of the solution to approximately 8.5 by adding ammonium
hydroxide. The solution should thicken as the pH is increased.

3. While stirring add 1 gram of hexafluoroacetylacetone (hfa) to the mixture.
Crystals should form immediately.

4. Empty the liquid and fluorescent crystals into a shallow container with a
large bottom surface and place on the heater table for a setting of '1' (70°C).

5. Allow the liquid to evaporate in a fume hood and thoroughly dry the crystals
(approximately 2hrs).

6. Using a razor blade, scrape the dried fluorescent precipitate off the bottom

and sides of the container and store in an appropriate container.

94

Fluorescent Paint Preparation:

The following is the procedure to prepare the fluorescent precipitate made in the

previous steps for application.

l.

10.

Weigh 10g of Methyl Ethyl Ketone (MEK) in a medium sized airbrush spray
bottle.

Add 0.15g of fine PMMA pellets to the MEK.

Place a magnetic stir bar in the bottle and cover the bottle with aluminum
foil.

Place the jar on the heat/stir table and set the heater at position '1' (70°C)
while stirring the solution.

Heat and stir the solution for about 30min or until the PMMA can no longer
be seen.

Turn off the heater and allow the solution to cool to room temperature while
stirring.

Weigh 0.05g of the fluorophore (Europium or Terbium) and mix the
solution.

Connect the bottle to the airbrush and adjust the flow of air as needed.

Apply a fine coat to the test surface and allow the liquid to evaporate. Care
should be taken not to over spray the surface since this will produce a non-
uniform polymer film.

Illuminate the sample with a handheld UV source to check the uniformity

and concentration of the fluorophore.

95

ll.

12.

Reapply coats as necessary until the desired effect is obtained. The polymer
film will not be evident to the eye.
Apply gentle heat to the coated surface to evaporate the MEK and 'set' the

polymer film. The film goes from transparent to a fine white fiosting.

96

Appendix B: Light Source Evaluation with Ocean Optics Spectrometer

Evaluation of the spectral output of the light sources used in this thesis was
accomplished with the Ocean Optics Spectrometer. A 3mm diameter fiber optic cable
(black), not originally included with the spectrometer system, was used to direct the
light output by the individual light sources into the spectrometer sample holder without
a sample present. The size of this fiber optic required removing the light collimator
screwed into the sample holder. A 600um diameter fiber optic (blue) included with the
spectrometer system was mounted 180° from the 3mm fiber. The opposite end of this
fiber was connected to the spectrometer. The other end of the 3mm fiber was held by
hand in front of the light source being evaluated until the spectrometer recorded the
spectral emission. Integration times were on the order of lOms for all the light sources
evaluated. Neutral density filters were placed in the filter slot located in the sample
holder in between the two fibers as needed to prevent the spectrometer from becoming
saturated due to the high intensities of the light sources. Typically, three low
transmissivity filters were needed to reduce the intensity of the light sources to a
measurable level. The evaluation of the light sources is presented in Figure B 2.

The Opti Quip light source is a high intensity Xenon source that produces
electromagnetic emission peaks at 365nm, 408nm, 439nm, 548nm, and 580nm.
Bandpass filters can be used to single out a particular peak for excitation of fluorescent
samples. A 365nm bandpass filter was mounted to the Opti Quip source for evaluation
of the F ERIT fluorophores. The evaluation of this configuration is denoted as "Opti

w/UV" in Figure B 2.

97

 

 

 

 

 

 

 

 

 

 

’6
d)
.5
E
o \
5
E
a:
c
.9.
E
200 300 400 500 600 700 800
Wavelength (nm)
—-—— Opti Quip — Opti w/ UV Ocean VIS
— Ocean W —— Dolan Jenner —— Sirchie UV-A

Figure B 2: Evaluation of the Spectral Emission of Electromagnetic Excitation
Sources with the Ocean Optics Spectrometer

The addition of the bandpass filter eliminated the other emission peaks included in the
original source.

The Ocean Optics visible and UV emission spectra are included in Figure B 2
and was discussed in the main part of the thesis. It should be noted that this source is
not as intense as the other sources evaluated. Integration times on the order of 605 were
necessary when using this source to evaluate Rhodamine B fluorescent samples. This
led to the use of the Dolan Jenner source, which reduced integration times to 10$ for the

same fluorophore concentration.

98

The Dolan Jenner source emits over the whole visible spectra with peak
emission near 600nm. However, a 488nm, lOnm bandwidth, filter was used with this
source to excite the Rhodamine B and Rhodamine 110 solutions.

The Sirchie UV—A source is a hand held light source that produces wavelengths
in the short UV wavelength region. This source was evaluated as an alternative to the

Opti Quip source to excite the FERIT fluorescent samples.

Appendix C: Matlab Code for FERIT and LIFI‘ Whole Field Analysis and AVI
Creation

The following Matlab codes convert gray-scale intensity images in tiff sequence format
obtained with the SensiCam digital camera and Image Pro Plus software into a pseudo
colored temperature avi file. The LIFT and F ERIT code differ in the equations used to
convert gray scale intensity ratios into temperatures. Besides that the codes are
relatively similar. These codes produced the images in Figure 3.10, Figure 3.12, and
Figure 3.15. The avi files produced by these codes are very large and should be

converted to mpeg format before using in a presentation.

LIFI‘ Code:

clear all

Ref = imread('ORge1ref.tif');

i=1;

y=1;

sumb=0;

for i=1 :5
B = imread('ORgelseq.tif‘,i);
sumb=double03)+sumb;

end

Bref=sumb./5;

norrnalizer = double(Bref)./double(Ref);

i=6;

while i<=995
B = imread('ORgelseq.tif‘,i);
ratio=double(B)./double(Ref)./normalizer;
temperature=-858.46.*ratio."5+2934.2.*ratio."4-3920.6.*ratio."3+2607.3.*ratio."2-

951 .85.*ratio+212.38;
imagesc(temperature,[20 80]); colormap(jet(15));
title('Oratec Monopolar Probe with Gel (30W, 75C set point)');
set(gca,'XTicklabel',{ },’YTicklabel',{ });
colorbar('vert');
tempmap(y) = getframe(gcf);
i=i+2;
FY“;

end

100

movie2avi(tempmap,'ORgel.avi',’quality', 100,'fps',1 5)

FERIT Code:

clear all

i=1;

y=l;

sumb=0;

for i=1 :1
B = imread('sp25eq.tif,i);
sumb=doublc(B)+sumb;

end

normalizer = sumb./1;

i=6;

while i<=50
B = imread('sp2seq.tif‘,i);
ratio=double(B)./normalizer;
temperature=107.241315030887+8.23585848058315.‘ratio-

848.008378147134.‘ratio."2+2358.20293067844.‘ratio."3-

2590.20299792734.*ratio."4+987.5232610519.*ratio."5;
imagesc(temperature,[20 90]); c010rmap(jet(20));
title('ArthroCare Bipolar Probe, (set point 2)');
set(gca,'XTicklabel', { },'YTicklabel',{ });
colorbar('vert');
tempmap(y) = getframe(gcf);
i=i+1;
y=y+1;

end

movie23vi(tempmap,'feritAC.avi','quality', 100,'fps',l 5)

101

BIBLIOGRAPHY

 

[2]

[31

[4]

[5]

[6]

[7]

[3]

[9]

[10]

[ll]

[12]

[13]

[14]

[15]

Neubert, P.,US Patent No. 2,071,471 (1937).

Amoczky, S.P., Aksan, A., 2001, ” Thermal Modification of Connective Tissues: Basic Science
Considerations and Clinical Implications,” American Academy of Orthopaedic Surgeons
Instructional Course Lectures, Sims, F.1-I. (ed.), AAOS, Chicago, IL, 50, pp.3-l 1.

Anderson, K., McCarty, EC, and Warren, RR, 1999, “Thermal Capsulorrhaphy: Where Are
We Today,” Sports Medicine and Arthroscopy Review, 7, pp. 117-127.

Perry, J.J., Higgins, L.D., 2000, “Anterior and Posterior Cruciate Ligament Rupture After
Thermal Treatment,” Arthroscopy, l6, 7, pp. 732-736.

Derby, R., et al., 2000, “Intradiscal Electrothermal Annuloplasty (IDET): A Novel Approach for
Treating Chronic Discogenic Back Pain,” Neuromodulation, 3, 2, pp. 82-88.

Saal, J.S., Saal, J.A., 2000, “Management of Chronic Discogenic Low Back Pain with a Thermal
Intradiscal Catheter - A Preliminary Report,” Spine, 25, 3, pp. 382-388.

Cilesiz, 1., Thomsen, S., Welch, A..l., 1997, “Controlled Temperature Tissue Fusion: Argon
Laser Welding of Rat Intestine In Vivo, Part One,” Lasers in Surgery and Medicine, 21, pp. 269-
277.

Brinkmann, R., et al., 2000, “Influence of Temperature and Time on Thermally Induced Forces
in Corneal Collagen and The Effect on Laser Therrnokeratoplasty,” Journal Of Cataract and
Refractive Smgery, 26, 5, pp. 744-754.

Kampmeier, J., et al., 2000, “Thermal and Biomechanical Parameters of Porcine Cornea,”
Cornea, l9, 3, pp. 355-363.

Edwards, RB, et al., 2002, “Thermal Chondroplasty of Chondromalacic Human Cartilage - An
Ex Vivo Comparison of Bipolar and Monopolar Radiofiequency Devices,” American Journal of
Sports Medicine, 30, 1, pp. 90-97.

Shellock, F.G., Shields, C.L., 2000, “Radiofrequency energy-induced heating of bovine articular
cartilage using a bipolar radiofiequency electrode,” American Journal of Sports Medicine, 28, 5,
pp. 720-724.

Kuo, T., etal., 1998, “Collagen Thermal Damage and Collagen Synthesis After Cutaneous Laser
Resurfacing,” Lasers in Surgery and Medicine, 23, pp. 66-71.

Fitzpatrick, R.E., et al., 1996, “Pulsed Carbon Dioxide Laser Resurfacing of Photoaged Facial
Skin,” Archives of Dermatology, 132, pp. 395-402.

Naseef III, 03., et al., 1997, “The Thermal Properties of Bovine Joint Capsule," American
Journal of Sports Medicine, 25, 5, pp. 670-674.

Hayashi, K., et al., 1996, “The Effect of Nonablative Laser Energy on the Ultrastructure of Joint
Capsular Collagen,” Arthroscopy, 12, 4, pp. 474-481.

102

 

[16]

[I7]

[13]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]
[23]
[29]
[30]

[31]

[32]

[33]
[34]

[351

Allain, J.C., et al., 1980, “Isometric Tensions Developed During the Hydrothermal Swelling of
Rat Skin,” Connective Tissue Research, 7, pp. 127-133.

Miles, C.A., et al., 1995, “The Kinetics of the Thermal Denaturation of Collagen in Unrestrained
Rat Tail Tendon Determined by Differential Scanning Calorimetry,” Journal of Molecular
Biology, 245, pp. 437-446.

Cilesiz, 1., et al., 1997, “Controlled Temperature Tissue Fusion: Argon Laser Welding of Rat
Intestine In Vivo, Part One,” Lasers in Surgery & Medicine, 21, pp. 269-277.

Pearce, J.A., et al., 1993, “Kinetics for Birefiingence Changes in Thermally Coagulated Rat Skin
Collagen,” Proceedings of SPIE, 1876, pp. 180-186.

Le Lous, M., et al., 1983, “Hydrothermal Isometric Tension Curves from Different Connective
Tissues. Role of Collagen Genetic Types and Noncollagenous Components,” Connective Tissue
Research, 11, pp. 199-206.

Kang, T., et al., 1995, “Heat-Induced Changes in the Mechanical Properties of Passive Coronary
Arteries,” Journal of Biomechanical Engineering, 117, pp. 86-93.

Privalov, P.L., 1982, “Stability of Proteins - Proteins Which do not Present a Single Cooperative
System,” Advances in Protein Chemistry, 35, pp.1-104.

Tang, J., et al., 1997, “Morphologic Changes in Collagen Fibers After 830 nm Diode Laser
Welding,” Lasers in Surgery and Medicine, 21, pp. 438-443.

Privalov, P.L., 1989, “Thermodynamic Problems of Protein Structure,” Annual Reviews in
Biophysics and Chemistry, 18, pp.47-69.

Olde Damink, L.H.H., et al., 1995, “Glutaraldehyde as a Crosslinking Agent for Collagen-Based
Biomaterials,” Journal of Materials Science: Materials in Medicine, 6, pp. 460-472.

LeCarpentier, G.L., 1993, “Continuous Wave Laser Ablation of Tissue: Analysis of Thermal and
Mechanical Events,” IEEE Transactions on Biomedical Engineering, 40, 2, pp. 188-199.

Omega Engineering, http://www.omega.com/techref/, 2002
Luxtron Corporation, http://www.luxtron.com, 2002

FLIR, http://www.tlirthermography.com/, 2002

Modest, M., F., Radiative Heat Transfer, McGraw Hill, Inc. 1993.

Lakowicz, J., R., Principles of Fluorescence Spectroscopy, (1999), Kluwer Academic/Plenum
Publishers, New York.

Allison, S., W., Gillies, G., T., ”Remote Thermometry with Thermographic Phosphors:
Instrumentation and Applications," Rev. Sci. Instrurn. 68 (7), July 1997. pp 2615-2650.

Rayleigh and Raman Scattering, Fluorescence and Phosphorescence, web reference.
Edwards, D.,l(., Radiation Heat Transfer Notes

Engler, RH, et al., "Pressure Sensitive Paint Systems for Pressure Distribution Measurements in
Wind Tunnels and Turbomachines,” Meas. Sci. Technol. (1 I), 2000. pp. 1077-1085.

1(I3

 

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

Allison,S.,W., Cates, M.R., Beshcars, D.L., "A Survey of Thermally Sensitive Phosphors for
Pressure Sensitive Paint Applications," Oak Ridge National Laboratory.

Lian, Bin, “Fluorescent Emission Ratio Imaging Thermography and its Application in Heat
Transfer Analysis,” PhD. Thesis, MSU, I999.

Edge, A.C., Laufer, G., Krauss, R.H., "Surface Temperature -Fie1d Imaging with Laser-Induced
Thermographic Phosphorescence," Applied Optics, vol.39, no. 4, Feb 2000. pp. 546-553.

Bhaumik, M.L., "Quenching and Temperature Dependence of Fluorescence in Rare-Earth
Chelates," the Journal of Chemical Physics, 10, (12), June 1964. pp. 3711-3715.

Ananias, Duane, et al., "Novel Microporous Europium and terbium Silicates," J. Am. Chem.
Soc., 2001, 123, 5735-5742.

Ranson, R.M., Thomas, C.B., Craven, M.R., "A Thin Film Coating for Phosphor
Thermography," Meas. Sci. Technol. 9 (1998). pp. 1947-1950.

Sakakibara, J ., Adrian, R.J., "Whole Field Measurement of Temperature in Water Using Two-
Color Laser Induced Fluorescence," Experiments in Fluids 26 (1999). pp. 7-15

Kim, H.J., Kihm, K.D., "Application of a Two-Color Laser Induced Fluorescence (LIF)
Technique for Temperature Mapping," 2001 ASME Intemation Mechanical Engineering
Congress and Exposition.

Aksan, A., Nielubowicz, D.S., McGrath, 1.1., 2001 “Modeling The Thermal Histories Of

Collagenous Tissues Subject To Different Heating Modalities,” HTD-24433, Proceedings of
ASME Winter Annual Meeting.

Cengel, Yunus, Heat Transfer, a Practical Approach, (I998), McGraw Hill, Boston, Mass.

104