Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
“CAT”-OLOGY
SPECTRALLY RESOLVED NEUROPHOTONICS IN THE MAMMALIAN BRAIN AND PHANTOM STUDIES
BY
KANDICE TANNER
B.S., South Carolina State University, 2002 M.S., University of Illinois at Urbana-Champaign, 2003
DISSERTATION Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Physics
in the Graduate College of the University of Illinois at Urbana-Champaign, 2006
Urbana, Illinois
ii
iv
Dedication
To my mother, Allisene,
And of course, Beryl the Cat.
v
Acknowledgements
Only by the grace of God have I made it this far, and his presence cannot be disputed.
Secondly, there are several people who have been instrumental in this process. First, I
must thank my advisor, Dr. Enrico Gratton, “Doc” who took me into his lab and accepted
me AS IS, and allowed a graduate student who was 99% wrong most days to relish my
days of being 1% right. The training and experiences cannot be easily expressed in
words. I am eternally grateful to him. I look forward to the day when I could finally beat
him at something.
My other advisor, Dr. Mantulin, “Dr. M”, Thank you for helping me in my research and
providing inspiration when I was in one of my many moods.
Julie Wright was like a mother and could take care of anything in the bat of an eye. I wish
she had moved to California.
To the people of the LFD, when I first joined the lab, how can you describe the
transformation from hell to utopia? I had a lot of fun and countless tea and chocolate
breaks. The LFD was a magical place. It is my sincere hope that Camelot will be realized
again in California.
Dr Monica Fabiani and Dr. Gabriele Gratton and members of the CNL who provided
invaluable help and limitless information about the physiology of the brain to a simple
physics graduate student.
Dr. Joseph Malpeli has been extremely helpful in all aspects to provide additional
training in areas where I had no knowledge. He also allowed me free access to his lab and
would meet with me to discuss my incessant questions. I would also like to thank Rui
Ma, graduate student in Neuroscience and official “cat handler” who tirelessly met with
vi
me every day for a month, including the entire thanksgiving break to acquire the last set
of data before the lab moved to California. I thank him for taking care of Beryl.
There are other people in the department who I must formally thank:
Dr Phillip Philips who took a chance on a student who knew the Quantum Hall
Coefficient at 3am. Who gives a physics exam at 3am? Dr. Phil.
Dr. Nathan who was extremely supportive throughout my career and for his honesty.
Wendy Wimmer has been like a system administrator, her efficiency and disposition
made it so much easier to deal with life’s little complications like what do I register for
again?
There are others too countless to mention who have helped from my schooling in
Trinidad to my matriculation in the US. Oh, I must acknowledge my funding source, the
NIH
Finally, to the future Dr. E. Red, who has been extremely supportive in this phase of
thesis writing I would have not been able to finish without him.
vii
Preface (Disclaimer)
My hope is to detail the journey that has brought me to this point of PhDom. In order to
write this thesis, I had to amuse myself in some form or fashion to complete it in a timely
manner. I certainly didn’t enter the field because I was an accomplished author. Hence, I
wrote it the same way I approach life, in that it is always better to laugh than to cry.
Also, I hope that the next Graduate student who enters the lab would find it useful, in the
very least as good bedtime reading who knows it may become a classic- essential reading
to cure insomnia, maybe my committee members would also enjoy this added benefit.
viii
Table of Contents
I. Introduction.................................................................................................................1
I.A Optical characteristics of tissue in the NIR- Absorption and Scattering................3 I.B Light Transport in tissue.........................................................................................7 I.C Theoretical Modeling- Multi-distance Frequency Domain.................................. 11 I.D Chapter Summary .................................................................................................13
II. Theoretical Modeling -Introduction of the Spectral Approach.................................14
II.A Introduction...........................................................................................................14 II.B Spectral Approach.................................................................................................14 II.C Applications of the Spectral Approach .................................................................15 II.D Chapter Summary .................................................................................................16
III. Investigation of Quasi-diffusive regime in Self –Reflectance geometry.................17
III.A Chapter Introduction.............................................................................................17 III.B Description of Self Reflectance Geometry...........................................................17 III.C Experimental Procedure .......................................................................................19 III.D Results ..................................................................................................................20 III.E Discussion ............................................................................................................24 III.F Chapter summary ..................................................................................................25
IV. Investigation of the Cat Visual Cortex- optical BOLD signal ..................................27
IV.A Introduction ..........................................................................................................27 IV.B Experimental Protocol ..........................................................................................27 IV.C Data Acquisition...................................................................................................31 IV.D Results ..................................................................................................................34 IV.E Discussion ............................................................................................................40 IV.F Chapter Summary.................................................................................................43
V. Phantom Studies –Modeling Vasodilation................................................................44
V.A Introduction ..........................................................................................................44 V.B Physiology ............................................................................................................45 V.C Model of Optical Properties of Tissue .................................................................45 V.D Experimental Procedure .......................................................................................46 V.E Data Analysis .......................................................................................................49 V.F Results ..................................................................................................................50 V.G Discussion ............................................................................................................51 V.H Chapter Summary.................................................................................................55
VI. Phantom Studies- Validating technique to recover known optical properties..........57
VI.A Introduction ..........................................................................................................57 VI.B Method..................................................................................................................57 VI.C Data Analysis .......................................................................................................58 VI.D Results ..................................................................................................................59
ix
VI.E Discussion ............................................................................................................61 VI.F Chapter Summary.................................................................................................62
VII. Investigation of Cat Visual Cortex- Fast Signal .......................................................64
VII.A Introduction ..........................................................................................................64 VII.B Physiology ............................................................................................................64 VII.C Experimental Procedure .......................................................................................65 VII.D Data Acquisition...................................................................................................67 VII.E Data Analysis .......................................................................................................70 VII.F Results ..................................................................................................................70 VII.G Discussion ............................................................................................................89 VII.H Chapter Summary.................................................................................................92
VIII. Investigation of Cat Visual Cortex-Pulse .................................................................93
VIII.A Introduction ..........................................................................................................93 VIII.B Simulations...........................................................................................................93 VIII.C Experimental Procedure .......................................................................................96 VIII.D Data Analysis .......................................................................................................97 VIII.E Results ..................................................................................................................99 VIII.F Discussion ..........................................................................................................105 VIII.G Chapter summary ...............................................................................................106
IX. Summary .................................................................................................................108 X. Future plans.............................................................................................................116
References........................................................................................................................117 Author’s Biography .........................................................................................................127
1
I. Introduction Physicists have made significant contributions to the field of biology and medicine. For
example as recently as 2003, the Nobel Prize for Physiology and Medicine was jointly
awarded to a physicist and a chemist for pioneering work in Magnetic Resonance
Imaging and its applications. This technique was only possible because of the work done
in the field of Nuclear Magnetic Resonance by solid state physicists in the 1940s. This is
just one example where medicine has benefited greatly from applications that were
originally non medical in nature. Another example is the discovery of X-rays by
Roentgen in 1895 is heavily exploited today as a non-invasive method of probing the
human body. The list is extensive. Currently, several groups are heavily involved in the
development of techniques, instrumentation and the basic comprehension of tissue
hemodynamics using visible and near Infrared (NIR) light.1-3
Figure 1.1 Schematic showing the electromagnetic spectrum and the regions in which different techniques are employed.
2
In Figure 1.1, we can see the electromagnetic spectrum and the range of wavelengths in
which different techniques are employed. If we consider this broadband spectrum,
different regions are widely used to probe biological tissue such as X-rays (0.1-10 nm),
Visible light (400-700nm) and Near-Infrared (NIR 650-1100nm). These techniques are
extremely beneficial as they can generate non-invasive, rapid, relatively high resolution,
functional images and real time devices. This in turn can aid the medical practitioners in
diagnosing strokes, hematomas, vascular deficiencies, tumors.4-6 However, these
techniques are not restricted to diagnostics but have great potential for treatment as seen
in recent studies of photodynamic therapy for cancerous tissue. Secondly, the information
that is obtained can aid in the understanding of muscle and brain physiology. The only
way that these optical techniques can be beneficial to clinicians is if one can obtain
quantitative results. However, this poses a severe problem as one must be able to describe
the transport of light in tissue where the light has traveled in the tissue and more
importantly understand processes such as scattering and absorption at both the
macroscopic and microscopic level. Additionally, there are several models that are
touted for light transport depending on the type of tissue, but one must choose one that is
compatible with in vivo measurement of the desired biological system. There is a
common misconception that in the field of biophysics, the concepts involve a hybrid of
interdisciplinary fields as opposed to the traditional probing of basic science. However, it
is quite the opposite as several basic principles of physics must be examined and
understood before we can understand the biological processes. I believe that the new
breed of physicists must also become experts in other disciplines where they may have
not had formal training. This work involved understanding the basic physics as well as
3
the application of these theories to describe physiological processes. The importance of
this study can be placed in two categories: one where the emphasis is on the information
that is gained in studying the basic physics of light-matter interactions, specifically
probing tissue in vivo and the other, the information that can be determined about
physiological processes that can be used for diagnostics, treatment or basic
comprehension of the complex system that is the human brain.
I.A Optical characteristics of tissue in the NIR- Absorption and Scattering
First, let us address each of the terms:
• Absorption
When atoms or molecules absorb light, the incoming energy excites a quantized structure
to a higher energy level where the ground state has the lowest energy. It must be noted
that transitions between states are only possible if the energy of the photon is comparable
to the energy differences between states. This process describes Absorption. It is useful to
think of these transitions by examining movement up the rungs of a ladder where it is
possible only to go from rung to rung, but not permissible to place your feet in between
rungs (at least if you have any intention of staying on the ladder).
The absorption spectrum of an atom or molecule depends on its energy level structure.
Absorption spectra are useful for identifying compounds. The energy of any molecule
can be described by the following equation:
Emol = Etrans + E e-spin + Enuc spin + E rot + E vib + Eelec (1)
where the contributions are due to the motion of the center of mass, electronic and
nuclear spin, rotation, vibration and electronic configuration respectively. Table 1 shows
the energy in eV that causes transitions in the respective energy states.
4
Type Frequency Magnitude (eV/molecule)
Translation Continuous Very Small
Spin RF 10-7
Rotational Microwave 10-3
Vibrational Infrared 0.1
Electronic UV/Visible 1-10
Table 1- Table displaying types of transitions and the energy and EM region associated with them
The energy (proportional to the inverse of the wavelength) of the incoming light
determines the information that can be gleaned from specific tissue. It is mainly
determined by the absorption spectrum (absorption of light as a function of wavelength)
of the illuminated tissue that is related to its energy states. Photons that correspond to the
X-Ray are the most energetic, and typically these energies are greater than the energy
spacing in the lighter biological molecules in tissue. Hence, the greatest contrast in tissue
imaging is seen with the bones as they contain heavier atoms like Calcium (greater
difference in energy states) where the energetic X ray photon can knock an electron from
the atom and is comparable to the energy differences. Visible light is more comparable to
the energy difference in soft tissue and causes transitions between the electronic states in
soft tissue. NIR causes transitions in the Vibrational states as shown in Figure 1.2.
5
Figure 1.2-Infrared Absorption In tissue, deoxyhemoglobin (HHb) and oxyhemoglobin (O2Hb) are the main absorbers in
the NIR (650- 1150 nm), accounting for 90 % of the photon absorption. 7-8 Melanin also
has a high absorption coefficient in this spectral window but as this chromophore is
restricted to the relatively thin superficial layers of tissue and hemoglobin comprises a
volume fraction ranging from 1-5%, the overall effect of melanin is negligible. Figure 1.3
shows the absorption spectra of tissue chromophores in the NIR.
Abs
orpt
ion
coef
ficie
nt: O
2Hb,
( m
m-1
mM
-1),
H2O
(mm
-1),
Mel
anin
(rel
ativ
e)A
bsor
ptio
n co
effic
ient
: O2H
b, (
mm
-1m
M-1
), H
2O (m
m-1
), M
elan
in (r
elat
ive)
Abs
orpt
ion
coef
ficie
nt: O
2Hb,
( m
m-1
mM
-1),
H2O
(mm
-1),
Mel
anin
(rel
ativ
e)
Figure 1.3-Therapeutic Window and Absorption spectra for the Tissue Chromophores. HHb and
O2Hb- are in units of mm-1mM-1 while water and fat- mm-1. The scattering spectrum in the figure
follows the λ-4 dependence
6
• Scattering
Scattering is primarily governed by the wavelength of the incident light, the size and
shape of the particle (radius), geometries, heterogeneities and the mismatch of refractive
indices within the observed medium. The distribution of sizes of particles as well as
random thermal fluctuations results in a non-uniform spatial and temporal distribution of
refractive indices. It results in the re-direction of light due to its interaction with matter
via the principle that the incident electromagnetic (EM) wave causes the oscillation of
electric charges and/or the excitation of the vibrational modes of the excited states. The
scattered or re-directed light is the relaxation of these vibrational modes or the
acceleration of these charges. Scattering may or may not occur with transfer of energy,
i.e., the scattered radiation has a slightly different or the same wavelength. There are
different types of scattering such as Rayleigh or Elastic scattering i.e. no energy loss (or
light in has the same frequency as the light out) and its angular distribution is symmetric.
This scattering also has a dependence on wavelength, specifically (λ-4). (Figure1.3).
Stokes Raman and Anti-Stokes Raman scattering are types of inelastic scattering where
the wavelength of the scattered light is larger and smaller respectively than the incident
wave. Mie scattering is dependent on the size of the scattering centers where the angular
distribution of the scattered light is highly forward. In tissue, the contributors for
scattering are the cells, cellular organelles, proteins resulting in the general heterogeneous
nature of tissue. Figure 1.4 (http://omlc.ogi.edu/classroom/ece532/class3/scatterers.html)
shows the hierarchy of the size distribution of tissue components.9 Rayleigh scattering is
primarily seen due to light interaction with membranes and macromolecular aggregates
on the order of 0.01-0.1µm, while Mie scattering is seen where the interaction with
7
mitochondria and vesicles on the order of a micron. Hence, in tissue there is a mixture of
Mie and Rayleigh scattering. Also most scattering in tissue is forward scattering.
Figure 1.4- Hierarchy of tissue components showing the type of scattering that is comparable.
I.B Light Transport in tissue
The modeling of light transport in tissue is dependent on the region of the EM spectrum
that is used for illumination as the relative magnitudes of the scattering with respect to
the absorption differ for each region. However, in all regions scattering and absorption
occur simultaneously and the main problem is to separate the two processes. This is not a
unique problem and has been explored extensively by Astrophysicists to describe the
process of the propagation of radiation through an atmosphere which is itself emitting
radiation, absorbing radiation and scattering radiation. Chandrasekhar’s Equation of
Radiative Transfer can be applied to stellar systems as well as that of light transport
through tissue; the underlying principles are the same as they simply involve how energy
is attenuated due to losses via absorption and inelastic scattering as well as the redirection
of light due to elastic scattering processes.10 Specifically, there are four independent
macroscopic parameters that are used to characterize light propagation in tissue: the
scattering anisotropy, (g), the absorption coefficient, µa, the scattering coefficient, µs , and
8
the index of refraction, (n). The definitions can be found in Table 2 as well as figure 1.5.
The reduced scattering coefficient, µs’ is a quantity that is given by the following
equation:
µs’= µs( 1-g) (2)
This can be used as a length scale for isotropic scattering events where the photon loses
all memory of its initial direction.11 These parameters give us sufficient information to
interpret the biochemical and structural properties of the investigated tissue. Also, the
penetration depth, l, of photons in tissue can be calculated by the following equation:12
l= (µs + µa)-1 (3)
• absorption coefficient, µa(cm-1 )-The inverse of the absorption mean free path.
• scattering coefficient, µs(cm-1 )-The inverse of the single-scattering mean free path.
• reduced scattering coefficient, µs’(cm-1 )-Approximate inverse scale of isotropic scattering.
Figure 1.5 Schematic showing definitions of absorption and scattering mean free paths
Scatter in g C en ters
A bsorber
Source D etector
Scatter in g C en ters
A bsorber
Source D etectorSource D etector
9
Parameter Definition
absorption coefficient, µa
(cm-1 )
The inverse of the absorption mean free
path
scattering coefficient, µs
(cm-1 )
The inverse of the single-scattering mean
free path.
reduced scattering coefficient, µs’
(cm-1 )
Approximate inverse scale of isotropic
scattering.
scattering anisotropy, (g) Average cosine of the scattering angle
index of refraction, (n) Ratio of speed of light in vacuum to the
speed in the medium.
Table 2- Description of terminology associated with tissue optics
Second, let us examine the different light transport modalities. We revisit the concept that
as light passes through tissue, some of it is scattered, absorbed and reflected. When
scattering is the dominant process, the light can be modeled as if it propagates in
spherical waves. This transport regime is known as the diffusive regime. See figure 1.6.
Detector fiber
Laser diodes
ReflectedCollected
Absorbed
Light
Diagram- Diffusive Regime
Figure 1.6- Diagram showing multi-distance technique for light propagation in tissue in diffusive regime.
10
However, for smaller distances and/or less-scattering tissue, the photons are not fully
randomized before they are detected, the quasi-diffusive regime. Lastly, when there is no
scattering, as in an optically clear medium, the light is just attenuated by the absorbing
chromophores but there is no change from its initial trajectory. However, the last case is
never observed in the case of tissue. A good analogy is to think of a game of bowling
where the ball represents a photon. In the case where we have no scattering, and if the
person is horrible at the game, a “gutter” ball one where the ball would go through
without hitting any of the bowling pins before exiting. Now, if the ball hits a few pins
before exiting the forward direction of the ball has not been affected by the interaction
with the pins, this is the case of the quasi- diffusive regime where scattering is introduced
but the photon has not been fully randomized. Now, if you can imagine a game where the
pins have an appreciable mass compared to that of the ball that results in the ball being
redirected at each collision in a total randomization before exiting, in fact it may even
come back in the direction in which it came initially, this is the diffusive regime. Figure
1.7 shows this concept in transillumination geometry.
Figure 1.7- Transillumination geometry showing different light transport regimes.
Optically Clear
Intermediate OpticallyTurbid
Non-Diffusive
DiffusiveIntermediate
Optically Clear
Intermediate OpticallyTurbid
Non-Diffusive
DiffusiveIntermediate
11
I.C Theoretical Modeling- Multi-distance Frequency Domain
Now, I have provided a qualitative description of light transport in tissue however, if we
wish to do a more quantitative analysis, we must consider the frequency of the incident
light. The following theoretical assumptions hold:
• µa >> µs’ , Beer- Lambert Law, λ < 300nm and λ > 2000nm
• µa << µs’ , Diffusion Approximation, 650 nm < λ < 1150 nm
• µa ~ µs’ , Monte Carlo and Equation of Radiative Transfer, 300nm < λ < 650 nm
and 1150nm < λ < 2000 nm
In the near infrared region, the absorption of the hemoglobin is greatly attenuated with
respect to the value in the visible region, such that scattering is the dominant process. For
example, the reduced scattering coefficient µs’ of the gray matter in the human brain
ranges from 20-30 cm-1 while the absorption coefficient µa is about 0.25cm-1.13 (Figure
3). This region is called the “therapeutic window” and the modeling of light transport for
this region is known as photon migration. The photons are randomized due to the
multiple scattering events in tissue in this spectral window. This allows us to use the
diffusion approximation to the Boltzmann transport equation:
(4)
• ψ-the fluence rate(W/m2)
• D -the reduced scattering coefficient = 1/ (µabs + µs)
• µabs - the absorption coefficient
• S is the light source ( isotropic)
There are physical constraints to this approximation such as the mean free path of the
scattering is much smaller than that due to absorption, the medium is homogeneous and
that the light source must be isotropic. Intuitively, this is an erroneous assumption as
( ) ( ) ( ) ( ) ( ) ( ) ( )kkabskk srtStrrtrrDtrtc
−=+∇⋅∇−∂∂ δψµψψ ,,,1
12
tissue by definition is heterogeneous. Also as advertised, scattering in tissue is mainly in
the forward direction, however on the length scale typical of the reduced scattering co-
efficient; the photon density can be assumed to be uniform.14,15,16 Hence, the restrictions
for this equation impose that the source-detector distance must be at least 1-2 cm and that
the photons have traveled at least one mean free path. Therefore, all measurements
cannot be near sources and boundaries. In summary, NIR provides sufficient penetration
in tissue and a light transport modality can be implemented to recover the absolute
concentration of the chromophores that are involved in physiological processes.
However, typically in the field two wavelengths are used to probe the concentrations of
O2Hb and HHb simultaneously. In the Gratton photon migration group, the Multi-
Distance Frequency Domain technique has been developed using amplitude modulated
light at moderately high frequency in the range (100-400MHZ) to study the
hemodynamics of the muscle and brain non-invasively. It has been well documented that
the frequency-domain approach of light propagation in tissues provides high temporal
analysis and a high signal to noise ratio.17,18 Scattering is separated from absorption by
examining the phase delay and attenuation of detected light as shown in figure 1.8.
Figure 1.8-Diagram showing the differences between the input and detected light.
13
I.D Chapter Summary
Several groups have been successful in separating the processes of absorption and
scattering in tissue to recover quantitatively the concentrations of O2Hb and HHb. In the
Gratton laboratory, the multi-distance Fd-NIRS method is used where the theoretical
model follows the diffusion approximation to the Boltzmann transport equation.
However, due to the physical limitation of the diffusion approximation and the
instrumental limitations of modulating more that two wavelengths simultaneously, the
need for a technique that enables us to have broadband access as well as be independent
of the light transport modality (diffusive or non-diffusive) is the next logical step in the
field of photon migration. This suggests that one must fully investigate the nature of light
transport in the quasi-diffusive regime, the true effects that the heterogeneous nature of
tissue has on the optical signals, and if the SNR and temporal resolution is comparable to
Fd –NIRS.
2/1
2
2
12
−
Φ
Φ⎟⎟⎠
⎞⎜⎜⎝
⎛+⋅⋅−=
DC
DCa S
SSS
νωµ a
a
DCs
Sµ−
µ=µ
3
2'
From (SDC, SΦ):
[ ] [ ]HbHbO HbHbOaλλλ εεµ += 22 [ ]
12
2
21
2
2
2
11
2
2
λλλλ
λλλλ
εεεεεµεµ
HbHbOHbHbO
HbOaHbOaHb−
−=
THC = [HbO2] + [Hb]
Semi-infinite homogeneous medium:ln(rDC) vs. r ln(rAC) vs. r Φ vs. r
SDC (µa, µs’) SAC (µa, µs’) SΦ (µa, µs’)
SaO2 = [HbO2] / THC
Two-wavelength multi-distance approach:
14
II. Theoretical Modeling -Introduction of the Spectral Approach
II. A Introduction
It was stated earlier that one of the major problems with the diffusion approximation was
that, to be able to recover accurate values, one must be in the diffusive regime typically
where the source-detector distances are greater than 1-2 cm (for the brain). The rule of
thumb says that in this regime, the penetration depth is equal to half the source-detector
regime. So in order to probe tissue depths of less than 5mm or where the photons have
not been fully randomized, we must employ a different technique. In this section, we
propose a new method which separates the absorption and scattering processes by
spectral analysis and is independent of the transport regime.
II. B Spectral Approach
In the spectral approach, we use a large number of wavelength points i.e. a broadband
spectrum; hence we can determine the absorbance of tissue components such as HHb,
O2Hb, fat and water and their spectral shifts with high precision. Assuming that we
know the spectrum of the individual tissue components, we can construct a linear
combination of basis spectra to fit the overall tissue absorbance spectrum (Equation 5).19
I = Scattering (λ) + Water (λ) + Fat (λ) + O2Hb (λ) + HHb (λ) + cytochrome (λ) (5)
The coefficients of each term as well as the scattering power n are determined using a
non-linear least squares method. By knowing the coefficients used to fit the absolute
spectrum, we estimate the fractional contribution of the individual components in the
measured tissue (figure 2.1). All data acquisition and analysis were performed by using
the Elantest software, originated in the LFD photon Migration group.
15
Figure 2.1- Comparison of experimental data with theoretical fit using the spectral approach In figure 2.1, one can see the experimental curve that is obtained from a human finger
and the theoretical fit using the spectral approach.
II. C Applications of the Spectral Approach
Experimental Outline
1) A reference spectrum is taken (lamp/baseline, I0).
2) Spectrum is collected (I) using Elantest and all measurements are made with respect
to reference [log (I/ I0)].
3) Weighted Components are determined using theory.
We apply the spectral approach to the differential spectrum obtained by subtracting the
baseline period from the stimulation period. The differential changes are small and we
assume that the changes observed can be described by a linear combination of the basis
components. However, for the application of the spectral approach in the differential
measurements for tissue, the wavelength dependence of the scattering is allowed to vary.
It is not strictly fixed to a λ-4 , which is invalid for tissue. Our system allows us to acquire
spectra at a frequency of 200Hz; hence, the relative tissue component contributions can
be determined with high temporal resolution. Consequently, we can see the optical
16
changes due to water, HHB, O2Hb and scattering as a function of time. These
measurements can be used to potentially detect the blood oxygenation level dependence
response (BOLD) due to the neuronal activation. The spectral method separates
scattering from absorption, as scattering has a characteristic spectral behavior, different
from any other spectral component. The spectral approach is different from the
“frequency domain” method that exploits the time of flight in the diffusive regime to
extract the scattering coefficient. One advantage of the spectral technique is that the
spectral shape (including the scattering contribution) is independent of the light transport
regime, i.e. it is applicable in both the diffusive and non-diffusive regime.
II. D Chapter Summary
This technique must be tested with phantoms and animal models to validate the claim that
this technique can be used to recover the optical signals accurately. Hence, we must apply
the technique to systems where we know the absorbing and scattering properties to
determine the accuracy. First, we must understand the quasi-diffusive regime. Second, we
need to establish if we have enough sensitivity using our instruments. Ultimately, we
would like to apply this technique to recover relative changes as those in physiological
studies as well as to recover absolute concentrations.
17
III. Investigation of Quasi-diffusive regime in Self –Reflectance geometry.
III. A Chapter Introduction
We model the different types of tissue based on their optical properties. The physical
structure of the human brain is sufficient to emphasize this point. The outermost portions
of the cerebral hemispheres are continuous folds of cortex comprising grey matter rich in
neuronal cell bodies and dendrites where most of the neuronal activity. This region is
intertwined with white matter rich in myelinated axons which account for its highly
reflective nature. The cerebral hemispheres are not uniform but are comprised of ridges
and grooves (gyri and sulci); hence, the thickness of grey mater with respect to the white
matter is not the same at each point.20 Additionally, for non-invasive measurements, light
must travel through the scalp, skull and Cerebrospinal Fluid (CSF) before encountering
the brain. If we consider the true geometry of the human brain and the optical properties
of its components, the question arises if the reflective white matter plays a critical role in
the observed optical signals of the human brain. Furthermore, is the effect independent of
the light transport regime? Elementary consideration from the basic laws of physics
shows that when light intersects a boundary with a mismatch in refractive indices, the
original trajectory of the light is altered. Hence, what remains is to quantify this effect in
the human brain.
III.B Description of Self-Reflectance Geometry
In our approach, we place our optodes in the side by side configuration known as the self-
reflectance mode where we model the tissue as if it were a semi-infinite medium. (Figure
3.1).
18
Figure 3.1- Panel a) Case of no scattering: left shows that the observation volume is small if the reflector surface is close to the surface. Only when the reflector surface is at a fixed depth with respect to the source-detector bundle, a large volume can be detected. Panel b) Case with scattering: left shows that scattering broadens the numerical aperture and the observation volume is larger at a smaller depth with respect to the source-detector bundle.
From figure 3.1, one can see that in this geometry the light that can be detected depends
on three parameters, one being the numerical aperture of the optical fibers, the position of
the reflecting boundary with respect to the optodes and finally the scattering properties of
the medium. The conical distribution of the fibers is given strictly by the formula,
numerical aperture, NA= sinα. Simply, light can only be detected when there is an
intersection of the dispersion volume (conic) of both source and detector. In panel a, one
can see that the closer the reflector is to the optodes (optical fibers for source and
detector), the less light can be detected and as this distance is increased more light can be
detected. Additionally, as S-D increases, the height increases as the conic volumes are
further apart and therefore intersection must occur at a greater height with respect to the
S S
S S
D D
D D
a)
b)
S S
S S
D D
D D
a)
b)Observation
Volume
19
reflector. As scattering is introduced, the conical volumes become distorted and the
intersection occurs at shorter distances.
III. C Experimental Procedure.
To test various aspects of our animal model, we also performed measurements on a
phantom. The phantom model we used is shown in figure 3.2.
Figure 3.2 Left shows the source- detector distances probed to go from non-diffusive to diffusive regimes, while the right shows the experimental setup The light source was a tungsten lamp at a nominal temperature of 3100K (LS-1 Tungsten
Halogen Light Source, Ocean Optics, Dunedin, FL, US). The spectrometer employed
was the model S2000, also from Ocean Optics. Both the light source and the
spectrometer were coupled with a 1000µm core diameter optical fiber. The phantom
consisted of a beaker filled with a milk solution to simulate the optical properties of the
skull and the gray matter. The bottom of the beaker was lined with white tape to simulate
the reflective nature of the white matter. The scattering properties of 2% milk are
comparable to that of the mammalian brain, where the reduced scattering coefficient, µs’ ,
of milk is 5cm-1 and the reduced scattering coefficient, µs’ ,of the brain is 5-10 cm-1 .13,21
The basic experiment consisted of determining the intensity of light as a function of
Source Detectors
4.5 6.0 8.7 9.6 mm Non diffusive--------------Diffusive
S-D
Tungsten Lamp
White Light Source
Data Acquisition
Ocean OpticsS2000 Spectrometer
(Detector)
4.5mm
BATH Height White Tape
20
height above the reflector using the Elantest software (This program is available at
ftp://www.lfd.uiuc.edu/lfd/egratton/elantest/) for each source detector configuration
denoted by “S-D” in figure 3.2, and for each milk solution. Four S-Ds were considered
4.5, 6.0, 8.7 and 9.6mm and for each “D” the height of the source-detector bundle was
controlled using a robotic arm that allowed the entire system to be moved in highly
accurate incremental steps of 0.625mm. The milk solutions investigated varied from
solutions with no scattering (water) to highly scattering (store bought 2% milk). The
notation used refers to the fraction of 2 % milk (store-bought) with respect to the total
solution (milk and water) and were as follows: 0 (pure water), 0.05, 0.1, 0.2, and 1,
where 1 correspond to “undiluted milk”.
III. D Results
As the source-detector distance, S-D, increased, one obvious observation was that the
maximum intensity attained decreased for all of the solutions. In addition, the height at
which the peak of intensity was seen increased as S-D increased. (Figure 3.3) This peak
was asymmetrical in each case. A second striking feature was the plateau that was
observed for the case of the scattering solutions. (Figure 3.3 b, c, d, and e) It was also
seen that as the scattering of the solution increased, the height at which this plateau
occurred decreased irrespective of “S-D”. In the extreme case, most scattering solution
(milk) showed that the plateau is observed after a few mm (Figure3.3e). In the case,
where there was no scattering (water), there was a trend to approach this plateau, but, the
full range of measurements needed to observe this effect was not examined. However, in
this case, it was clear that a significant light intensity is not observed until the height
above the reflector was at least 20mm. (Figure 3.3a) Further investigation of figures
21
3.3b,c and d show that the maximum intensity attained is greatest in the case where the
scattering is greatest, (0.2C > 0.1C > 0.05). From Figure 3.4a, c, e, g, the height at which
the maximum peak is attained is much greater in the case of the optically clear medium
(water) than for the turbid solutions for all Ds. Furthermore, intensities for the turbid
solutions were at their respective plateaus. Additional investigation of each D, for heights
up to 10mm, (figure 3.4b, d, and h) showed that the maximum intensity was always seen
for the 0.2C solution and the minimum for the water. However, the intermediate values
vary for each of the solutions considered.
22
Figure 3.3- Panels a-e show the intensity versus height for different milk solutions.
Intensity vs HeightWater
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height0.05C
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height0.1C
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height0.2C
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)In
tens
ity (c
ount
s)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height2% fat milk
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs HeightWater
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height0.05C
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height0.1C
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height0.2C
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)In
tens
ity (c
ount
s)
4.5 mm6.0 mm8.7 mm9.6 mm
Intensity vs Height2% fat milk
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
4.5 mm6.0 mm8.7 mm9.6 mm
b)
c) d)
e)
a)
23
Figure 3.4- Panels a- h show Intensity versus height as a function of source-detector distance and a detailed insert to show the heights corresponding to 0-10mm above the reflector.
Intensity vs HeightD=4.5mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=4.5mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=6.0mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=6.0mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=8.7mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=8.7mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=9.6mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=9.6mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=4.5mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=4.5mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=6.0mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=6.0mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=8.7mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=8.7mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=9.6mm
0
500
1000
1500
2000
2500
0 10 20 30 40 50 60
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
Intensity vs HeightD=9.6mm Detailed
0
500
1000
1500
0 2 4 6 8 10
Height (mm)
Inte
nsity
(cou
nts)
Water0.05C0.1C0.2CMilk
a b
c) d
e) f)
hg
24
III. E Discussion
These experiments clearly show the principle that the diffusive regime in the
geometry of the semi-infinite medium is a function of both the scattering properties of
the medium as well as the source-detector distance, S-D. If we consider the
experimental setup, we see clearly that for cases of no scattering, (optically clear), the
detected light is simply a function of the numerical aperture of the optical fibers and
the height with respect to the reflector surface (Figure 3.1a). Simply, light can only be
detected when there is an intersection of the dispersion volume (conic) of both source
and detector. This is illustrated experimentally in the case of the optically clear
solution (water) where we see that the light intensity measured remains negligible
until the height has reached at least 10mm. Additionally, as S-D increases, the height
increases as the conic volumes are further apart and therefore intersection must occur
at a greater height with respect to the reflector. This changes as we introduce
scattering into the solution, as the conic volume of the optical fibers become
distorted. (Figure 3.1b) Hence, in the case of the strongly scattering medium (milk),
after only a few mm, the photons detected increase dramatically. But, the maximum
value attained will be less than in the optically clear solution as some photons will be
lost. However, for the intermediate solutions (0.05, 0.1 and 0.2), the maximum
intensity is not a function of the scattering. This fits with the model as in the quasi-
diffusive regime, the photons are not completely randomized, hence the detected light
can either increase or decrease for a specific S-D. Another striking feature of these
graphs is that despite increasing the height after a certain value, the intensity remains
the same, (plateau). This shows that after these heights the detected intensity is
25
independent of the reflector and we are in the diffusive regime. Hence, these graphs
clearly show the transition from the quasi-diffusive regime to the diffusive regime. It
also shows that in the case where the optical properties were similar to that of the
human brain, (2% milk), we see that the reflector (white matter) restricts the number
of photons detected only for heights of a few mm. Hence, confirming the results as
reported by Okada et al.22However, for practical experiments where the light has not
been fully randomized as in the quasi-diffusive regime, we see that the intensity can
either increase or decrease even if the scattering increases for a fixed source-detector
distance at heights less than 12mm with respect to the white matter. It must be stated
that in this experiment the source –detector geometry remained the same. However, if
the source and detector were interchanged with respect to each other, a different result
would be observed in the quasi-diffusive regime as we cannot say that the light
detected is equivalent to the symmetrical “banana bundle” observed in the diffusive
regime as the photons are not fully randomized before they are detected. This is
another concern that must be accounted for when working in the quasi-diffusive
regime.
III. F Chapter Summary
The results are important for several reasons. The true nature of tissue can have a
significant effect on the observed optical signals; hence we must understand the type and
size of the effect of the different types of tissue under investigation. It is clear that for the
probing of the brain, not only is the source detector separation important to determine
depth of tissue penetration but the position of the white matter with respect to the grey
matter. Additionally, these experiments provided information about the signal to noise
26
ratio when in the quasi-diffusive regime. These results gave a preview into the expected
results of the in vivo measurements by first examining a phantom model. Specifically, we
can now be sure that we have a high SNR to recover the true optical properties in the
animal model.
27
IV. Investigation of the Cat Visual Cortex- optical BOLD signal
IV. A Introduction
In this chapter, we present the experimental results of the animal model studies. First, we
must explain why we chose to perform experiments using the visual cortex of the cat. The
questions that must be addressed are if the technique is independent of the light transport
modality, if it can recover the appropriate optical signals which result from physiological
changes and if our current instrumentation has a comparable SNR to the frequency
domain instrumentation. The cat provided us with an excellent positive control for the
following reasons: extensive knowledge of the cat’s visual cortex was developed by one
of our collaborators (Dr. Joseph Malpeli). Specifically, he determined the exact location
of the neuronal activation based on the type of visual stimulus presented to the cat. The
anatomy of the cat is such that the photons that enter the tissue are not fully randomized
before detection in the reflectance geometry as the thickness of the tissue probed was
3mm where the thickness of the skull was 1mm, and the grey matter was 2mm. Hence,
we are not in the diffusive regime. In addition, the physiological changes are expected to
be similar to that of humans. Hence, it is ideal to test if new technique yields results that
are comparable to Fd-NIRS such as Blood Oxygenation Level Dependence (BOLD)
effect.
IV. B Experimental Procedure
Methods- Cat protocol
1. Preparation of the cat
These experiments were performed on one adult female cat. All procedures were in
accordance with U.S. Public Health Service Policy and protocols approved by the
28
University of Illinois Institutional Animal Care and Use Committee. All surgery was
performed aseptically and under general anesthesia. A gold plated ring was implanted
under the conjunctiva of one eye to allow eye movements to be monitored using the
double magnetic induction method.23 The scalp and muscle overlying the calvarium were
removed, and an aluminum fixture surrounding this area, bonded to the skull to provide
support for a protective cap that was also used to immobilize the cat’s head during the
experiment. A mixture of clear acrylic cement and antibiotics ~ 1 mm thick was then
placed in lieu of the removed tissue to provide a permanent protective barrier.24 Metal
tubes (15 gauge, thin wall hypodermic tubing, inner diameter, 1.52 mm and outer
diameter, 1.83 mm), 7 mm in height, were then embedded in this mixture above the
region corresponding to the visual cortex, area17/18 and the motor cortex, area 4. Figure
4.1 shows schematically the location of the metal tubes.25 The tubes were in direct
contact with the acrylic mixture above the bone. For simplicity, a grid system was
implemented to indicate the position of each tube; this is used to identify the source and
detector locations. The “a” row was located above the frontal lobes and each tube was
numbered in sequence from 1-5. The “b” row was located parasagittally, roughly over
the border between areas 17 and 18 of the visual cortex in the right hemisphere. The “c”
row was located in a coronal plane that cut across the region of area 17 near the surface
of the brain, as well as across the entire coronal extent of areas 18 and 19 of the visual
cortex (figure 4.1). The intersection of the “b” and “c” rows is roughly at the center of the
representation of the area centralis. The distance between adjacent tubes was 2mm, with
the exception of c4 and c5, where the “c” row intersects the “b” row. The tubes served as
holders for the source and detector optical fibers during our measurements.
29
a1………..a5
b1……
.....…b7
c1……………...c7
Figure 4.1- Atlas of the cat’s brains showing the placement of the tubes for optical fibers
2. Visual Stimulation and Behavioral Paradigm
The cat was positioned facing a rear projection screen subtending 60o horizontally and
50o vertically at a distance of 70 cm from the animal’s eyes. The screen was illuminated
uniformly at 0.021cdm-2 with white light from an LCD projector. A computer controlled
laser spot approximately 0.1o in diameter served as a fixation point on the center of the
screen. Visual stimuli consisted of periodic flashes generated by white LED clusters that
were superimposed on the screen raising the luminance to 2.020cdm-2 during the flash.
The cat sat immobile in a bag with its head fixed to a rigid plate, tilted forward 5o with
respect to the Horsley-Clark horizontal plane. During a trial, the cat was trained to focus
on the fixation point regardless of the flashes which had no behavioral significance.
Generally, if the cat maintained this fixation (within +/- 7o) for 10 seconds, it was
rewarded with food. The inter-trial interval was 10 seconds. However, during data
collection sessions, the cat was rewarded at the end of each trial regardless of its
performance. For the purposes of this experiment there were two types of tasks, one in
which the cat was visually stimulated (VS trial), and one where there was no visual
stimulation (NVS).
30
1. VS trials consisted of 10 seconds of a repetitive sequence of 20 flashes where the
flash was on for 250 ms and off for 250 ms, terminated by the reward followed by a 10
second inter-trial interval.
2. NVS trials consisted of 10 seconds of no flashes, with delivery of the reward
followed by a 10 second inter-trial interval.
The trials were done with the NVS trials performed first (in blocks of 100) followed
immediately by the VS trials (with minimum perturbation). This was achieved by
toggling an external switch to activate the flash. Neither the cat nor the optical setup was
disturbed in any way. The cat was monitored at all times with an infrared sensitive
camera and light sources. The detector tubes were largely shielded from these light
sources.
Experimental Procedure
1. Technical Aspects
Spectral measurements were performed using an Ocean Optics (830 Douglas Avenue,
Dunedin, FL 34698, USA) detector system consisting of a S2000 spectrometer, an
ADC2000 PCI card and a tungsten lamp. The tungsten lamp gives a continuum spectrum
following Planck’s blackbody spectrum at a temperature of 3100 Kelvin. The PCI
ADC2000 hardware interface between the spectrometer and the computer performed an
analog to digital conversion at a sampling frequency of 2 MHZ at a 12 bit resolution
which allows spectral acquisition every 5 milliseconds. Additionally, free running
operations and external trigger modes were available for synchronizing external events.
The S2000 Ocean Optics is a miniature spectrometer with large spectral response (350 –
1100 nm) and good spectral resolution (0.3 – 10 nm). The spectral response is optimized
31
for the NIR range. This was achieved by a combination of different accessories: a
diffraction grating with a spectral response in the 550-1100 nm range (#4 in Ocean Optics
catalog) and a long pass filter to transmit wavelengths greater than 550 nm. These
combined with the response of the CCD linear array detector, pre-mounted on the
spectrometer, gave us the required spectral range of analysis in the NIR region, 650- 990
nm. The optical resolution and spectral response depend on the slit entrance on the
spectrometer, groove density of the gratings, fiber optics diameter and number of
elements (pixels) of the detector. Optimization of light detection through highly
scattering tissues was achieved by using an entrance slit of 200 microns and a fiber optic
core of 1000 microns. The grating had a groove density of 600 mm-1 and the CCD array
has 2048 pixels. The combinations of these parameters gave us an optical resolution of
4nm.
IV.C Data Acquisition
The cat was immobilized in a bag facing the screen with its head fixed to ensure minimal
movement. The optical fibers were then positioned in the tubes on the cat’s head. Any
particular pair is indicated according to the labeling system used in figure 4.1. For
example, a configuration of b4b6 refers to the placement of the source fiber in tube b4
and the detector fiber in tube b6. The tip of each fiber was placed in direct contact with
the acrylic which was roughly 1-2 mm above the cat’s skull, and approximately 4mm
above the surface of the cat’s brain. The cat’s head was held rigidly fixed during the
experiment, once the tubes were placed in the tubes; consequently their positions were
stable and could be reproduced from session to session. The spectrometer was armed at
the beginning of each trial by an external pulse coincident with a short auditory signal to
32
indicate the start of the trial. Data acquisition and analysis were performed by the
Elantest software (This program is available at
ftp://www.lfd.uiuc.edu/lfd/egratton/elantest/), which communicated through the PCI
ADC2000 card with the S2000 spectrometer. Synchronization of the data acquisition
with the visual stimulation was achieved by an external Stanford Research (1290-D
Reamwood Avenue, Sunnyvale, CA 94089, USA) pulse generator (model DS345). The
trigger and synchronization system provided the correct time signals to the Elantest
software and the external trigger input port of the spectrometer. Communication with the
software via the parallel port of the PC enabled the system for the start of data
acquisition. The spectrometer acquired spectra every 5 ms. First, a reference spectrum
was taken under the condition of no trial, meaning that the cat observed the screen at its
normal background level as described in the previous section, but otherwise not
performing either task. All differential measurements were calculated with respect to this
initial spectrum. Equal blocks of data consisting of 100 trials were collected under
different conditions. Data acquisition began with the beginning of each trial. The
collection of spectra was synchronized with externally supplied pulses that were
coincident with the onset of each flash in the VS condition (i.e. every 500ms). The same
timing was also provided for the NVS conditions. The spectrometer was armed at the
start of each trial by an initial pulse and acquired a spectrum every 5 ms, giving a total of
95 spectra corresponding to 475 ms. The spectrometer then waited for the next trigger,
which occurred 500 ms after the start of the trial. This cycle was repeated 16 times for a
total of ~16* 95 spectra corresponding to about 8 seconds of the trial. No spectra were
acquired during the final two seconds of the trial, the reward and the inter-trial interval.
33
Each spectrum had 2048 wavelength points. The data matrix was then saved and the
sequence repeated 100 times. The procedure was repeated for different source- detector
configurations. It was identical for NVS trials except that the pulses provided to
synchronize the collection of spectra every 500ms were not accompanied by flashes. The
experiment was performed with minimum perturbation as the only change during the data
acquisition was the activation of the flashes during the VS trials via an external computer.
(See figure 4.2).
Figure 4.2- Schematic showing the synchronization achieved using an external system, and the placement of the source-detector fibers while the cat is observing the screen.
3. Data Analysis
In this chapter, we present the analysis of the hemodynamic signal. The analysis of the
fast neuronal signal will be discussed in a separate chapter. Data were collected for the
first 8 seconds of each trial. However, every 475 ms, one spectrum was deleted due to
the external synchronization of the system, as this spectrum had a different integration
time (30 ms instead of 5 ms). This spectrum was disregarded, but the overall time axis
was maintained correctly, i.e., the matrix had a gap of 30 ms every 500 ms. Since we
were using this sequence only for slow signals, and the time axis was correct, this
Start acquisition pulse
500 ms250 ms
25 ms95 pulses
Visual stimulation signal
Pulse sequence
t
0
0SC
REEN
S D
Start acquisition pulse
500 ms250 ms
25 ms95 pulses
Visual stimulation signal
Pulse sequence
t
0
0SC
REEN
S D
500 ms250 ms
25 ms95 pulses
Visual stimulation signal
Pulse sequence
t
0
0SC
REEN
S D
500 ms250 ms
25 ms95 pulses
Visual stimulation signal
Pulse sequence
t
0
0SC
REEN
S D
34
deletion had no influence on the final result. The remaining spectra were then folded
across the 16 flash cycles and averaged over the desired number of trials. Folding was
also done as a function of the number of trials to see if there were any differences in the
signals observed due to physiological changes occurring across trials. A definitive
spectral pattern and intensity pattern changes were observed in the raw data matrix. We
performed spectral deconvolution and principal component analysis (PCA) on the raw,
but folded data. From the PCA, it was determined that the minimum number of basis
components required to correctly fit the differential spectrum was four: scattering, water,
HHb and O2Hb. The spectrum was then separated into the weighted contributions of
these individual species and their changes observed as a function of time. No
assumptions were made with respect to the baseline hemoglobin concentrations as we
only consider a differential spectrum. In representing the changes of the components as a
function of time, only O2Hb and HHB were plotted using the same scale. The scale of
the changes in water component and that of the scattering component are arbitrary.
IV. D Results
For all of the different source-detector (S-D) configurations, data were processed as
described in the previous section to detect both the changes due to the hemodynamic
signal, as well as to the fast (millisecond) neuronal signal. However, due to the volume
of information that was obtained, only certain aspects of the hemodynamic signal will be
presented in this chapter. First, we present maps of the raw data matrix after folding
followed by the separation of this matrix into the weighted contributions of scattering,
water, O2Hb, and HHb as functions of time. The maps show the average change for all
of the wavelengths as a function of time for the folded data. Finally a comparison of the
35
optical BOLD effect seen as a function of source–detector (S-D) configuration is shown.
Although we show results for only a few experiments, the results are highly reproducible
for any given S-D combinations. For locations in the visual cortex, the optical BOLD
effect is observed during visual stimulation and not seen in the absence of the
stimulation. Similarly for S-D configurations located outside of the visual cortex (frontal
lobes), the Optical BOLD effect was not observed for NVS or VS trials.
1. Raw Data
Maps displaying the raw data matrix of each S-D pair over the visual cortex, areas 17 and
18 show a significant change in the shorter wavelengths for the first 3 seconds of the
visual stimulation. (Figures 4.3a, 4.3c). The data matrix for the control configuration that
was positioned over the frontal lobes, where we didn’t anticipate any changes, remains
relatively constant during the same activation period. (Figure 4.3e). For all of the
configurations during NVS trials, the changes in wavelengths as a function of time were
also minimal (Figures 4.3b, 4.3d and 4.3f).
36
Figure 4.3- Raw Data Matrices showing the relative OD as a function of time and wavelength. 2. Spectral Deconvolution
The raw data matrix was decomposed for each time bin (linked to trial length) using 4
spectral components: O2Hb, HHb, water and scattering. Figures 4.4a-4.4f show three
different S-D locations for VS and NVS trials, one from the para-sagittal row of tubes
over the area 17/18 border (b4b6; figure 4.4a, 4.4b), one roughly in a coronal plane over
areas 17 and 18 (b4c6; figure 4.4c, 4.4d) and one over the frontal lobes (a2a4; figure 4.4e,
4.4f). There is an optically detected BOLD effect due to the activation of the visual
cortex in which the O2Hb starts to increase after 1-2.5 seconds after visual stimulation.
37
This delay is dependent on the area investigated. Differences in the values and time
courses of the signals due to changes in O2Hb and HHb are detected between regions in
the visual cortex which differ by about 2mm. Additionally, significant changes are seen
due to scattering and water. The change due to scattering also has a different time course
with respect to the other components. For source–detector (S-D) configurations that
correspond to areas in the visual cortex, a maximum decrease (trough) in scattering is
seen at roughly the same time as the O2Hb reaches a maximum in the optical BOLD
effect during VS trials. The water component is seen to decrease at the onset of
stimulation and tracks the HHb changes with some time delay (figures 4.4a, 4.4b).
Control experiments (figure 4.4e, 4.4f) show that there are no major changes in the O2Hb
and HHb (BOLD effect) that track with the visual stimuli for the S-D pair located over
the frontal lobes, and the contributions from these components are one order of
magnitude smaller than the signal obtained from the VS trials (Figures 4.4a, 4.4c). The
signals due to the changes in water and scattering are on the same order of magnitude but
they don’t appear to be correlated to visual stimulation. In addition, control experiments
show minimal signal changes in the absence of visual stimulation (figure 4.4b, 4.4d,
4.4f).
38
0 2000 4000 6000
-0.05
0.00
Rel
ativ
e O
D
Time/ms
b4b6 VS(a)
0 2000 4000 6000
-0.05
0.00
Scattering Water Oxy Deoxy
Time/ms
b4b6 NVS(b)
0 2000 4000 6000
-0.02
-0.01
0.00
0.01
Rel
ativ
e O
D
Time/ms
b4c6 VS(c)
0 2000 4000 6000
-0.02
-0.01
0.00
0.01
Scattering Water Oxy Deoxy
Time/ms
b4c6 NVS(d)
0 2000 4000 6000
-0.01
0.00
0.01
0.02
Rel
ativ
e O
D
Time/ms
a2a4 VS(e)
0 2000 4000 6000
-0.01
0.00
0.01
0.02
Scattering Water Oxy Deoxy
Time/ms
a2a4 NVS(f)
Figure 4.4-Spectral Deconvolution showing changes in scattering, water, O2Hb and HHb as a function of time.
39
3. Optical BOLD effect as a function of S-D locations.
Figures 4.5a and 4.5b show the optically detected BOLD effect for several S-D locations,
two parallel to the 17/18 border (b4b6, b3b5), and one S-D pair located over the frontal
lobes (a2a4), and three that roughly straddled this border orthogonally (b4c6, b4c5,
c5b5), respectively. The signals for the two S-D configurations parallel to the 17/18
border show similar temporal behavior. However, for the configuration b4b6 the
amplitude of the signal is much larger (about a factor of 4) than for any other
configuration. For the orthogonal S-D pairs, it was observed that the amplitude of the
signal due to the optical BOLD effect was approximately the same. However, there were
different delays among the different pairs. In one configuration, b4c6, an initial dip is
seen lasting for ~2 seconds. The initial dip was not observed at all locations, indicating
that its magnitude is not constant everywhere.
0 2000 4000 6000
-0.005
0.000
0.005
0.010
0.015
O2Hb b4b6 HHb b4b6 O2Hb b3b5 HHb b3b5 O2Hb a2a4 HHb a2a4
Abs
orpt
ion
Time/ms
BOLD EffectParasagittal, roughly parallel to the 17/18 border
(a)
0 2000 4000 6000-0.005
0.000
0.005 O2Hb b4c5 HHb b4c5 O2Hb b4c6 HHb b4c6 O2Hb c5b5 HHb c5b5
Abs
orpt
ion
Time/ms
BOLD Effect orthogonal to the 17/18 border
(b)
Figure 4.5-Comparison of O2Hb and HHb as a function of cortical position
40
IV. E Discussion
This work is the first to determine changes of the spectral components in the mammalian
(cat) brain due to external visual stimuli. Previous work has focused on the
determination of the reflectance signal in the near-IR with the purpose of imaging
columnar neuronal organization26, and the optically detected BOLD effect was not
determined in those studies. Furthermore, the fast dynamics were obtained using long-
integration, high sensitivity cameras in which the acquisition was triggered at different
times after visual stimulation rather than utilizing rapid spectral acquisition which was
synchronous with the visual stimulation as in the present approach. The spatial resolution
that is obtained with our instrument is on the order of millimeters, which when compared
to that obtained using multi- distance NIRS in human subjects (centimeters) still gives us
a relatively high spatial resolution. One concern is that the radiation emitted by the
tungsten lamp will be unstable as it is temperature dependent. However, before any
analysis was done, the data was block averaged and folded across trials as is standard in
similar studies. Hence, any random fluctuations that are present would be removed before
the spectral deconvolution is performed. We have observed changes in the O2Hb, and
HHb signals during visual stimulation reminiscent of the fMRI BOLD effect. The origin
of the BOLD effect is attributed to an increase in blood flow following neuronal activity
in one part of the brain. This causes a decrease of the HHb level because HHb is washed
out. In the fMRI signal, only the decrease of the HHb concentration is measured,
whereas with the optical technique, we have access to both O2Hb and HHb. Upon visual
stimulation, we can clearly see the decrease in the HHb and the quasi simultaneous
increase in the O2Hb. Careful examination of the data in figures 4.5a, 4.5b shows that
41
there is a delay between the two signals. A similar delay between the increase of the
O2Hb and decrease of HHB was also observed in optical experiments in humans and was
attributed to oxygen consumption.27, 28 A striking feature of our experiments is that the
optical BOLD effect starts to decrease after a few seconds, (figure 4.5a) although the
stimulation continued throughout the trial.
Another novel outcome of these experiments is that the water signal, as well as the
scattering signal, changes during stimulation. The changes in the water component are
quite surprising since we expect no net change of the tissue water content during the short
time of visual stimulation. For the water content in the tissue to change rapidly (in
seconds), the water should be replaced by something else, which is not plausible. To
explain the decrease of the water content during the optical BOLD effect we must
consider the origin of the optical signal in the tissue. The tissue is highly heterogeneous
both from the physiological and the optical point of view. There are tissue regions that
are optically opaque, such as relatively large blood vessels. If, as we expect, these
vessels change diameter to accommodate an increase of blood flow following neuronal
stimulation, then this opaque optical compartment will increase. The net effect of the
increase in the opaque compartment results in an effective decrease of all the spectral
components, including the O2Hb and HHB. However, these tissue chromophores
undergo additional physiological changes due to the exchange of O2 and CO2 in the
tissue, which results in a net increase of the O2Hb signal (washout effect). We propose
that this optical effect caused by the presence of the opaque blood vessels is at the origin
of the apparent decrease of the water component. Following this reasoning, the water
component could be used to assess the extent of vasodilation due to stimulation. With
42
regard to the scattering signal, it should also decrease due to this optical effect. However,
there are also changes in the size of the microcapillaries, larger blood vessels and
possible changes at the cellular and sub cellular level. Therefore, the scattering signal
will not necessarily track the water signal.
All data presented in figures 4.5a-b refer only to the first 10 trials of a series of 100 trials.
As the trials continued there was a gradual change of the shape of the optical BOLD
effect. This fatigue effect will be discussed separately.
We observed that the optical BOLD effect is strongly dependent on the location in the
visual cortex. Locations 2 mm apart gave significant differences both in the amplitude
and in the time course of the signals.
We have observed several differences between the optical BOLD effect for the cat, when
compared with similar experiments in humans. It is premature to discuss the differences,
when only one animal has been studied. Furthermore, we cannot state unequivocally that
these results are “cat invariant”. However, additional cats are being trained to repeat the
experiments, and to assess if there are differences among individual animals. Certainly
the size of the cat’s brain could generate hemodynamic responses which differ from those
in humans.
One consideration for this discussion is whether or not the spectral approach could be
extended to studies of the human brain. Firstly, the size of the skull will bring the regime
of light propagation into the diffusive regime. However, the spectral approach is
independent of the modality of light propagation. There are two additional factors that
must be considered for the application of this spectral approach in humans. First, with
the present experimental apparatus, the S/N is insufficient to observe the dynamic
43
changes of the tissue chromophores at the distance of centimeters. For the slow
hemodynamic signal, using more sensitive detectors and more efficient monochromators,
we should have enough light for an accurate determination of spectral components.
Secondly, for the fast signal due to neuronal activation, where a very short integration
time is needed, the amount of light needed could be critical. Hence, additional work is
needed for the improvement of the technical aspects of this work.
IV. F Chapter Summary
In summary, we have developed a technique that is independent of the light modality
(diffusive or non-diffusive), where a broad-band spectral approach is used to determine
the individual NIR spectrum of tissue components. For the application in a mammalian
brain, we have examined the behavior of the scattering, O2Hb and HHB (BOLD effect)
simultaneously with other tissue components such as water content. The technique has
proven to have high enough temporal and spatial resolution to adequately determine the
localized hemodynamics. The behavior of water during stimulation has not been
discussed in previous literature. Our proposed model satisfactorily accounts for the
apparent change in the water content which can be used to better qualify the role of water
in vascular dynamics.
44
V. F Phantom Studies - Modeling Vasodilation
V. A Introduction
In the previous chapter, we determined that there was an apparent decrease of water
concentration and scattering concomitant with the optical BOLD effect in an animal
model.19 In regions of the cat visual cortex, we found that as a direct result of visual
stimulation, an optical BOLD effect (increase of O2Hb concentration and decrease of the
HHb concentration) was observed. However, the change in water concentration and
decrease of the scattering contribution were unexpected in the cat model. Previous
measurements of brain activation in humans did not have access to measurements in a
broad spectral range. This paper reported for the first time that there was an apparent
change in water concentration in tissue that occurred during the course of brain
stimulation that lasted for several seconds. Physiologically, this rapid change (in seconds)
of water content cannot be readily explained. It is highly improbable that there would be
such a large change (1-2%) in water concentration on the time scale considered in our
experiments. We proposed that the apparent water change could be due to an optical
artifact. In chapter V, we focus on this surprising result. We reason that vasodilation is
associated with brain stimulation and that vasodilation could be the origin of this optical
artifact. We simulated vasodilation with phantoms, where the water content was kept
constant. Our goal with the phantom studies is to demonstrate that simulated vasodilation
decreases all spectral components of a mixture. Another purpose of this model is to
estimate the amount of spectral changes due to “vasodilation” when compared to the
absorption of the tissue chromophores and scattering.
45
V. B Physiology
Neurovascular coupling or functional hyperemia in the mammalian brain relates the
vascular response to an increase in metabolic rate.29 In simple terms, in response to local
neuronal activity, vasodilation occurs and provides an increased supply of nutrient rich
blood to the activated region. Physiological processes result in exchange of the O2Hb
(oxyhemoglobin) with HHb (deoxy-hemoglobin) such that there is an increase in the
O2Hb and a quasi-simultaneous decrease in HHb during stimulation; the classic BOLD
(Blood Oxygenation Level Dependence) effect as seen using fMRI and optical techniques
by many authors.
V. C Model of Optical Properties of Tissue
The basic idea of the phantom model is that there are at least two optical compartments in
the brain tissue; one opaque, which is comprised of the large (larger than millimeter)
blood vessels and a second compartment, the brain tissue, which includes the small
capillaries (less than 100 microns) (Figure 5.1).
A B
S D DS
A B
S D DS
Figure 5.1- Schematic showing that the size of the opaque vessels change as a function of time will have an effect on the observation volume.
46
We define the opaque compartment as the contribution of those structures that appears
“black” at all wavelengths. Due to the physiological response to external stimuli, the
opaque compartment increases in volume as the result of vasodilation, thereby increasing
the total absorption (at all wavelengths). However, the relative amount of the
chromophores in the tissue decreases. Consequently, vasodilation causes an apparent
decrease in the absorption of all spectral components, including O2Hb, HHb, water and
scattering by the same relative amount (Figure 5.2).
Decrease in relative OD o f all components
Increase in background OD
Figure 5.2-Schematis showing the effect that vasodilation has on Absorption spectra
In the brain tissue, the increased blood flow due to vasodilation causes exchange of HHb
with O2Hb. Therefore, we should observe an increase for the O2Hb spectral component.
V. D Experimental Procedure
The phantom model was set up as shown in Figure 5.3, where the source-detector
distance was 4.5mm. The spoke assembly and the source detector were immersed in a
bath of milk (store bought 2% milk). The scattering properties of 2% milk are
47
comparable to that of the mammalian brain, where the scattering coefficient, µs , of milk
is 5cm-1 and the scattering coefficient, µs ,of the brain is 5-10 cm-1 .13,21 The light source
was a tungsten light source at a nominal temperature of 3100K (LS-1 Tungsten Halogen
Light Source, Ocean Optics, Dunedin, FL, US). The spectrometer employed was the
model S2000, also from Ocean Optics. Both the light source and the spectrometer were
coupled with a 1000µm core diameter optical fiber. The spectrum of the light source was
recorded prior to each measurement using a reflector in place of the milk solution. The
basic phantom design uses a pinwheel comprising of spokes that were rotated by a
stepper motor at a rate of 0.167 rev/s in the path of the “light bundle”. We examined a
specific condition where the size and position of the opaque objects was varied while
keeping the geometry of the optical setup fixed. A reference spectrum was taken of the
“bath”; namely, the milk solution in which the pinwheel rotated. Spectra were taken at a
rate of 50 spectra/sec while the spokes were rotated, which was comparable to the timing
of the spectra acquisition (200 spectra/sec) used in the cat experiments as reported.19
48
Figure 5.3- Schematic showing the Experimental setup for the phantom studies
1. Case I- Black Spokes of varying diameters
The pinwheel was comprised of four solid black spokes of diameters: 0.9mm, 1.2mm,
2.3mm and 3.5mm. These sizes were chosen to be comparable to that of blood vessels.
The spokes mimicked the effect that a change in diameter would impose on the optical
signals observed as the diameters of blood vessels are on the order of a few mm to cm.30
The entire system was placed at a depth of 1mm with respect to the source- detector
configuration. The depth of the pinwheel was then varied using a robotic arm that
allowed the entire system to be moved in highly accurate incremental steps of 0.625mm
and the entire procedure repeated.
Computer: Data acquisition and Analysis
Tungsten Lamp White Light
Source
Ocean Optics S2000
Spectrometer (Detector)
4.5mm
BATH
1mm
49
V. E Data analysis
Data acquisition and analysis were performed by the Elantest software created by E.
Gratton. (This program is available at ftp://www.lfd.uiuc.edu/lfd/egratton/elantest/). First,
a reference spectrum of the “bath” was taken. All measurements were calculated with
respect to this initial spectrum, i.e., only the changes with respect to the spectrum of the
milk bath are reported. The overall spectral changes were calculated by averaging the
changes at all wavelengths. The method of spectral deconvolution and data manipulation
was previously discussed in the animal study in extensive detail. In short, we determined
the changes of the tissue chromophores, including scattering, by forming a linear
combination of the spectrum of each individual tissue chromophore. For case I a
spectrum of the milk solution was also loaded as a basis spectrum to be used in spectral
deconvolution. Relative changes in spectral components are reported in terms of average
contribution to the total (average) absorption. Principle Component Analysis (PCA) was
used to determine the minimum number of spectra required to fit the differential
spectrum. In case I, the milk spectrum was sufficient. The spectrum was then separated
into the weighted contributions of these individual species and their changes observed as
a function of time as the spokes rotated under the source-detector pair. The relative
changes of these components as a function of time were plotted. In addition, the values of
the relative contribution to the absorbance were then plotted to show the relationship
between the relative absorption and spoke diameter (case I)
50
V. F Results
Case I- Black Spokes of varying diameters
As the opaque spoke intercepts the light bundle there is an increase of the total light
absorption (Figure 5.4a). The absorption change is uniform at all wavelengths.
However, when spectral deconvolution is performed for the “milk” component and the
total spectral changes are reported as a function of time, there is a relative decrease of the
amount of “milk” measured when the spoke is in the light path (Figure 5.4b). This effect
increases in a non-linear fashion as the spoke diameter increases (Figure 5.4c). As the
depth of the pinwheel was increased with respect to the source-detector position, one sees
that the effect of the opaque object is still appreciable up to a depth of 6mm (Figure
5.4d).
51
c)
y = 0.035x2 - 0.2645x + 0.158R2 = 0.9989-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
00.5 1.5 2.5 3.5 4.5
Diameter (mm)
Cha
nges
in A
bsor
banc
e
1
1.62
5
2.25
2.87
5
3.5
4.12
5
4.75
5.37
5 6
0.9 mm1.2 m
m2.3 m
m3.5 m
m
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
Cha
nges
in A
bsor
banc
e
Height (mm) Diam
eter
d)
0.9 mm
1.2 mm
2.3 mm
3.5 mm
Figure 5.4- Panel a) Raw data as the black spokes rotate, b) Spectral deconvolution into basis spectrum of the milk, c) The mathematical dependence of the changes in absorbance with spoke diameter, d) Changes in absorbance as a function of diameter of spokes and height below source-detector bundle.
V. G Discussion
The resolution of spectral components in a mixture is a classical problem in spectroscopy.
A common approach is the establishment of the number of independent components
using methods such as principal component analysis.31 When the components of the
mixture are known a priori, then simple spectral deconvolution is sufficient to determine
their relative contributions based on the Beer-Lambert law. In the case of tissue
spectroscopy, there is the additional issue in regard to the validity of the Beer-Lambert
law. In general, in the presence of strong scattering, the Beer-Lambert law must be
52
modified to account for the changes in optical path introduced by multiple scattering. In
our work, we are mainly interested in relative changes of the amount of spectral
components due to physiological activity. We assume that we know the spectral
components and we want to determine their relative changes. What is unique of our work
is that we want to determine fast changes in the brain following a stimulus: from a few
milliseconds corresponding to neuronal activity to several seconds due to metabolic
processes. Furthermore, we need to interpret changes occurring in the brain of a
relatively small animal (the cat). Due to the small size of the cat brain, the total optical
path of the light in our configuration is not sufficient for achieving the conditions of the
diffusion approximation. Also, we believe that superficial effects have a much stronger
influence in our results than for the case of human studies. Spectroscopy of the open
brain in small animals has been studied before using reflected light.32 In our case, we use
fiber optics to illuminate a point on the skull and light is collected after traveling several
millimeters in the brain. In our experiments, light travels some distance in the tissue
before been collected by the detector fiber. In this research we study light propagation in
a phantom system that closely mimics the situation of the cat brain. The source detector
distances as well as the distance from the surface, where we put our scatterer or
absorbers, is similar to the known geometry of the cat brain. Several groups have
addressed the inhomogeneous nature of tissue in both the diffusive and in the cases of
small (a few mm) optoelectrode separation and explored the second order corrections that
must be made to existing models of tissue dynamics.33-42 Our results from the dynamic
phantom made of the various diameters of opaque spokes show that “vasodilation” per se,
as simulated by changing the diameter of the spokes, could cause an apparent decrease in
53
all spectral components, including water and scattering. Therefore in the presence of
vasodilation of the large blood vessels, all spectral components should decrease
proportionally to their contribution to the overall spectrum. We propose that the apparent
decrease in water content (and partially scattering) observed in the cat brain following
visual stimulation is in fact due to vasodilation. In this model, the time course of the
water change reflects the dynamical changes of the diameter of the blood vessels. We
also demonstrated that the quantitative relationship between apparent absorption changes
and vessel (spoke) diameter is non-linear for small diameters (for diameters up to
3.5mm). However, when the diameter of the blood vessel becomes small, the vessel
becomes transparent, in contrast to the spokes in the phantom, which are always opaque
at all diameters. Our phantom studies also show that there is an apparent decrease of the
scattering spectral component when vasodilation occurs. The interpretation of the
changes in scattering in brain studies should consider this effect. However, in the brain,
there is a possibility that there are additional “true” scattering changes due to variation of
the size or number of scattering centers. In the following discussion, we estimate the size
of the changes in spectral components due to vasodilation for comparison with the
changes in chromophore concentration reported in the literature during the optical BOLD
effect.19,43 Typical changes in the concentration of tissue chromophore (O2Hb and HHb)
during stimulation are on the order of 1-2% both for humans and for the cat model.19,43
Similar values are found for the changes in scattering coefficient. It has been estimated
that vasodilation changes the diameter of the blood vessels by about 20%.44 To use our
data to estimate the effect of vasodilation in the human brain, we should estimate the
average diameter of the large blood vessels and also their depth with respect to the
54
surface. From figure 5.4a we can estimate that for a vessel diameter of about 2 mm, a
change of 20% of the diameter will give a change of about 0.04 OD units. However, this
change will depend on the depth of the vessels. At a depth of 6mm or more, the changes
will be much smaller (figure 5.4b). Since the average apparent optical density in tissue is
about 1-2 units, for source-detector distances of a few millimeters, the relative changes
due to vasodilation should be about 4% for superficial blood vessels but less than 0.4%
for vessels at 6 mm or more from the surface. In the case of the cat brain, the changes
due to vasodilation could be more significant in the measurements due to the small size
of the brain with respect to humans.
Note, that this artifact in the estimation of spectral components due to vasodilation would
not have been recognized if we had only used a few wavelengths. In fact, using a broad
band spectral analysis allowed us to distinguish changes that equally affect all spectral
components from specific changes affecting only one spectral component. The effect of
vasodilation on the optical spectrum is shown schematically in figure 5.2. Vasodilation
decreases the total light transmission (offset in figure 5.2) and reduces the spectral
amplitude. Therefore, if we measure only relative spectral changes, we will measure a
reduction of the spectral amplitude at all wavelengths. There is nothing special about
vasodilation in regard to optical tissue spectroscopy. Other mechanisms that broaden the
relatively large blood vessels should produce a similar artifact. For example, the pulse
could have a similar effect. However, following the above discussion, the pulse due to
relatively deep blood vessels has a very minor effect on the overall spectrum offset. Only
pulsating arteries close to the skin surface should produce an appreciable effect in
reducing all spectral components. This observation could explain a common yet seldom
55
reported observation that the pulse is barely visible when measurements are obtained
from deep layers when using, for example, the frequency-domain multi-distance method.
Instead, any method based on steady state continuous illumination (CW measurements)
or time resolved measurements at one point should be affected by the “vasodilation”
artifact which is more prominent at the surface. Our experimental results using phantoms
are independent of the model of light propagation in tissues. They are simply a
consequence of the fact that we cannot “see” inside opaque compartments. Furthermore,
our results can have implications on the way one interprets changes in chromophore
components in the presence of vasodilation, or any other physiological condition that
changes the relative contribution of the “opaque” compartment.
V. Chapter Summary
Our phantom model is clearly a simplification of the real brain structure as we only
consider two compartments, while in the actual brain we presumably have a continuum of
vessel sizes. However, the expected response due to intermediate vessel size will be a
mixture of the two extreme cases (opaque vs. transparent). Even if the real brain
situation is more complex, our conclusion in regard to the decrease of the relative
contribution of the tissue chromophore due to vasodilation is still valid. Note that our
model applies both in the diffusive (human brain) or quasi-diffusive regime (cat brain), as
this effect is seen in the raw data and therefore is independent on the algorithm used for
analysis. When using a few wavelengths and a single measurement point, there is no
simple way of determining the origin of the apparent optical changes (vasodilation or true
spectral changes). As a consequence of our studies, we conclude that determination of
56
chromophore concentration in tissue using only few wavelengths (for O2Hb, HHb and
scattering) is not sufficient to characterize the origin of the changes.
57
VI. Phantom Studies- Validating technique to recover known optical properties
VI. A Introduction
In our second set of phantom studies, we had to prove that we can accurately recover the
changes due to scattering and absorption, separately. This demonstration is vital to
validate the spectral method. In our phantom, we mimic spectral changes in absorption
and scattering by inserting spokes which contain different concentrations of absorbers
and scatterers. Our experimental results using phantoms are independent of the model of
light propagation in tissues.
VI. B Method
Case II- Spokes with varying Scattering Coefficients
Six spokes were attached to the pinwheel and placed at a depth of 1mm with respect to
the source-detector configuration. The pinwheel was made of black rubber in which holes
were made in a radial manner to secure test tubes (wall-thickness 0.4mm) filled with
different scattering solutions. Each test tube was secured using Parafilm to ensure that
there was no leakage of the solutions in the test tubes into the bath. The solutions
consisted of different ratios of milk to water. The notation used refers to the fraction of 2
% milk (store-bought) with respect to the total solution (milk and water) and were as
follows: 0.05 (same as that of the “bath”), 0.1, 0.2, 0.333, 0.5 and 1, where 1 correspond
to “undiluted milk”. For the case II experiments, the bath was made of milk diluted
1/20.
Case III- Spokes with varying absorbing materials (Low Absorbance Range)
For this series of measurements the test tubes were filled with a mixture of the “bath”
solution (milk diluted 1/5) and Blue gel food coloring (sold under the brand name Betty
58
Crocker gel Food colors). The initial dye mixture was diluted by adding the milk
solution (at a dilution equal of that of the bath) in appropriate amounts to get the desired
concentrations. The concentration used is shown as a percent concentration of the blue
dye in the overall mixture as follows: 3.125%, 6.25%, 12.5%, 25%, 50% and 100 %. The
Absorbance of the 100 % mixture was determined using a spectrophotometer (Perkin-
Elmer Lambda 5, Shelton, CT USA) with a standard 1cm cuvette and it gave an
absorbance of 1.44 OD units at 650 nm. The absorption coefficient, µa , could not be
calculated for this dye as the molecular weight was not known. However, the dye was
purely absorbing with no scattering.
Case IV- Spokes with varying absorbing materials (High Absorbance Range)
For these measurements, four spokes were used and the concentrations were 75, 80, 90
and 100 % of a solution of the dye with an Absorbance of 3.3 OD units at 650 nm.
VI. C Data Analysis
Data acquisition and analysis were performed by the Elantest software created by E.
Gratton. (This program is available at ftp://www.lfd.uiuc.edu/lfd/egratton/elantest/). First,
a reference spectrum of the “bath” was taken. All measurements were calculated with
respect to this initial spectrum, i.e., only the changes with respect to the spectrum of the
milk bath are reported. The overall spectral changes were calculated by averaging the
changes at all wavelengths. The method of spectral deconvolution and data manipulation
was previously discussed in the animal study in extensive detail. For all cases a spectrum
of the milk solution was also loaded as a basis spectrum to be used in spectral
deconvolution. For the cases III and IV, in addition to this milk solution spectrum, the
spectrum of the blue dye obtained using a spectrophotometer was also added to the
59
library of basis spectra. Relative changes in spectral components are reported in terms of
average contribution to the total (average) absorption. Principle Component Analysis
(PCA) was used to determine the minimum number of spectra required to fit the
differential spectrum. The milk spectrum was sufficient in all cases, case II required, in
addition to the 0.2 milk solution spectrum, scattering and for the case of the cases III and
IV, the spectrum of the blue dye. The spectrum was then separated into the weighted
contributions of these individual species and their changes observed as a function of time
as the spokes rotated under the source-detector pair. The relative changes of these
components as a function of time were plotted. In addition, the values of the relative
contribution to the absorbance were then plotted to show the relationship between the
relative absorption as a function of concentration for cases II-IV.
VI. D Results
1. Case II- Spokes with varying Scattering Coefficients
As the scattering of each spoke increased, we measured an increase in the scattering
spectral component when spectral deconvolution was performed (Figure 6.1a). In the
case where the spoke contained the same solution as the bath (0.05 milk solution), only a
very small change in the scattering was observed, probably due to thickness of the walls
of the spoke (Figure 6.1a). Furthermore, the changes in the milk spectrum were
extremely small (two orders of magnitude less) in comparison to the changes in the
scattering signal (Figure 6.1a). Hence, we conclude that the changes due to the mismatch
of the refractive indices between the surfaces of the milk solution and the cuvette
produced a negligible effect. The plot of the recovered magnitude of the scattering
60
spectral component as a function of scatterer concentration shows a linear relationship at
low scatterer concentration (0.05-same as bath, 0.1, 0.2, and 0.33) and then displays
saturation as the scatterer concentration increases (0.5, 1-same as pure milk) (Figure 6.1a
and b) .
Figure 6.1 a) Spectral deconvolution into scattering and milk solution, b) Spectral deconvolution into blue dye and milk for low OD, c) Spectral deconvolution into blue dye and milk for high OD.
2. Cases III and IV- Spokes with varying absorbing materials (Low and High
Absorbances)
There is an appreciable change seen in the optical signal due to the addition of the blue
dye (low optical densities) when spectral deconvolution is performed (Figure 6.1b) using
a)
b) c)
a)
b) c)
61
as a spectral component the spectrum of the dye obtained in a clear (non-scattering)
solution. In this regime, a plot of the absorption of the blue dye spectral component as a
function of concentration shows a linear relationship (r2= 0.9976) (Figure 6.1b and figure
6.2c). In the case of the highly absorbing dye solution, a saturation of the recovered
amount of spectral component is seen (Figure 6.1c and figure 6.2d).
-0.4
0
0.4
0.8
1.2
1.6
2
0 0.2 0.4 0.6 0.8 1
Concentration
Cha
nges
in A
bsor
banc
e
R2 = 0.9976
0
0.2
0.4
0.6
0.8
0 20 40 60 80 100
% Concentration
Cha
nges
in A
bsor
banc
e
0.55
0.75
0.95
1.15
75 80 85 90 95 100
% Concentration
Cha
nges
in A
bsor
banc
e
R2 = 0.9758
-0.4
0
0.4
0.8
1.2
1.6
0.05 0.15 0.25 0.35
Concentration
Cha
nges
in A
bsor
banc
e
a) b)
c) d)
-0.4
0
0.4
0.8
1.2
1.6
2
0 0.2 0.4 0.6 0.8 1
Concentration
Cha
nges
in A
bsor
banc
e
R2 = 0.9976
0
0.2
0.4
0.6
0.8
0 20 40 60 80 100
% Concentration
Cha
nges
in A
bsor
banc
e
0.55
0.75
0.95
1.15
75 80 85 90 95 100
% Concentration
Cha
nges
in A
bsor
banc
e
R2 = 0.9758
-0.4
0
0.4
0.8
1.2
1.6
0.05 0.15 0.25 0.35
Concentration
Cha
nges
in A
bsor
banc
e
a) b)
c) d)
Figure 6.2- a) Case of low scattering, plot of linear relationship of changes in absorbance with concentration of scattering, b)Case of highly scattering solution showing plateau, c) Case of low OD showing linear relationship of changes in absorbance with concentration of blue dye, d) Case of high OD showing plateau.
VI. E Discussion
With respect to the true spectral changes (addition of absorbing material), we
demonstrated that the relative absorption changes are linear with the dye concentration
only at low absorbances. At high absorbances, the apparent changes saturate. This is also
62
true for classic Beer-Lambert with no scattering. We show that we can accurately recover
(in two distinct cases) the spectrum (which is measured independently using a
spectrophotometer) of the absorber and the spectral component due to scattering as a
function of concentration (of absorber and scatterer). The range of concentrations was
restricted to the regime of low absorbance. It is seen as advertised that in the diluted
absorber/small scattering changes regime, the relative changes and concentration of
absorbers/scatterer as recovered by the measurement method will follow a linear
relationship. When the concentration of the absorbers and scattering centers increases,
the optical changes are no longer proportional to the concentration of the
absorber/scatterer and the changes start to saturate the measured absorption. Hence, we
demonstrated that saturation of the absorption could occur in real samples and that the
result is to produce artifacts in optical measurements.
VI. F Chapter Summary
In our phantom, we mimic spectral changes in absorption and scattering by inserting
spokes which contain different concentrations of absorbers and scatterers. We show that
we can accurately recover (in two distinct cases) the spectrum (which is measured
independently using a spectrophotometer) of the absorber and the spectral component due
to scattering as a function of concentration (of absorber and scatterer). The range of
concentrations was restricted to the regime of low optical density. It was observed that in
the diluted absorber/small scattering changes regime, the relative changes and
concentration of absorbers/scatterer as recovered by the measurement method followed a
linear relationship. When the concentration of the absorbers and scattering centers
increases, the optical changes were no longer proportional to the concentration of the
63
absorber/scatterer and the changes start to saturate the measured absorption. Our results
can have implications on the way one interprets changes in chromophore components in
the presence of vasodilation, or any other physiological condition that changes the
relative contribution of the “opaque” compartment.
64
VII. Investigation of Cat Visual Cortex- Fast Signal
VII. A Introduction
In this chapter, we return to the animal studies to establish if our technique is sensitive to
the fast optical signal. At the conclusion of the last study, it was determined that the
signal to noise ratio achieved with the old experimental equipment was insufficient to
convincingly detect this fast signal even though the optical BOLD effect was clearly
observed. We acquired new more sensitive equipment as well as changed the
experimental protocol to maximize the SNR for the neuronal response. We observed
some rather startling results where the heartbeat was clearly seen in each trial as well as a
fast global hemodynamic response on the order of 300 ms that has not been described in
the literature. Additionally, we have observed a fast signal with a rise time of less than
100ms that has no wavelength dependence which has been described in the literature as
the fast neuronal signal due to the changes in scattering in areas of the brain that deal
with the somatosensory cortical responses. More importantly, the observed signals are
seen in as little as one trial and observed using a purely spectrally resolved method. In
addition, we observe the highly debated “initial dip”. Here, I discuss the physiology of
this event and focus on this novel observation of the fast optical signal that is due to
stimulation of somatosensory cortex.
VII. B Physiology
In the brain, there are two responses to stimuli as reported in the literature using several
techniques including NMR, magnetic and electrical recordings and optical techniques.45-
51 These signals, fast and slow, are classified by the time they are seen from the onset of
stimulation as well as the physiological origin of the signal. In the context of optical
65
detection, the fast signal, usually on the order of milliseconds after the onset of stimulus,
is said to be seen as a change in the scattering signal due to the contraction of the neurons
during stimulation as they undergo a voltage change as associated with their action
potentials.52-56 This signal has been extensively described in the literature in experiments
in isolated tissue slices and more recently in the exposed cortex of live animals.57,58 The
slow signal, which has also been described previously, is seen seconds after the onset of
stimulation and is the direct result of the coupling of neuronal activation with the local
hemodynamic increase that arises from the vessels dilating to provide oxygenated blood
to the area of activation. There is a great debate about the very initial phase of the
hemodynamic response known as the initial dip.59,60 Physiologically, this initial dip
describes the rapid initial increase in HHb with the quasi-simultaneous decrease of O2Hb
at the onset of stimulus in the activated brain cortex.
VII. C Experimental Procedure
Methods- Cat Protocol
The same cat was used for these experiments hence the preparation of the animal as
described previously is applicable for this study.
Visual Stimulation and Behavioral Paradigm
The cat was positioned facing a rear projection screen subtending 60o horizontally and
50o vertically at a distance of 70 cm from the animal’s eyes. Visual stimuli consisted of
periodic flashes generated by white LED clusters that were superimposed on the screen
raising the luminance to 2.020cdm-2 during the flash. The cat sat immobile in a bag with
its head fixed to a rigid plate, tilted forward 5o with respect to the Horsley-Clark
horizontal plane. During a trial, the cat was trained to focus on the fixation point( a red
66
laser point) regardless of the flashes which had no behavioral significance. Generally, if
the cat maintained this fixation (within +/- 7o) for 10 seconds, it was rewarded with food.
The inter-trial interval was 15 seconds. However, for our experiments, during data
collection sessions, the cat was rewarded at the end of each trial regardless of its
performance. For the purposes of our experiment there were two types of tasks, one in
which the cat was visually stimulated (VS trial), and one where there was no visual
stimulation (NVS).
1. VS trials consisted of 10 seconds of a repetitive sequence of flashes terminated by
the reward followed by a 15 second inter-trial interval.
2. NVS trials consisted of 10 seconds of no flashes, with delivery of the reward
followed by a 15 second inter-trial interval.
For these experiments, we did two sets of trials, where each set consisted of one hundred
trials for the two conditions of VS and NVS to give us a total of two hundred trials. The
trials were then repeated for different stimulation frequencies and for different source-
detector positions. The number of flashes in the VS trials was varied corresponding to a
frequency of 1 Hz, 2 Hz, 2.5Hz, 4 Hz and 5 Hz, in that there were 10 flashes for the 1 Hz
(500ms on, 500ms off), 20 flashes for the 2Hz (250ms 0n, 250 ms off), 25 flashes for the
2.5 Hz (200 ms on, 200 ms off), 40 flashes for the 4 Hz (125 on, 125 off) and 50 flashes
for the 5Hz ( 100ms on, 100ms off).
The trials were done with the VS trials performed first (in blocks of 100) followed
immediately by the NVS trials (with minimum perturbation). This was achieved by
toggling an external switch to activate the flash. Neither the cat nor the optical setup was
disturbed in any way. Subsequently, to better understand the origin of the hemodynamic
67
signal, control experiments were performed where the visual paradigm remained the
same (flashes on and off) but the reward was not given to the cat.
Technical Aspects
For these experiments, the equipment used comprised of two HR (High Resolution) 4000
spectrometers obtained from the Ocean Optics (830 Douglas Avenue, Dunedin, FL
34698, USA) and a tungsten lamp coupled with 1000µm optical fibers to each of the
spectrometers (detector fibers) and the tungsten lamp (source fiber). The spectrometers
were sensitive to the NIR for the range of 680-1100nm, which is different to the
spectrometer in the first study where the spectrometer was sensitive for the range of 650-
990nm. Hence, it was more sensitive to the higher wavelengths which allowed us to
investigate the signal due to changes in the water band with a better SNR than in the
previous study. Secondly, the spectrometers were able to acquire spectra at a rate of
1ms/spectrum but we maintained the previous rate of sampling at 5ms/spectrum. Also
these spectrometers allow rapid data transfer to the host computer via an USB II
interface.
VII. D Data Acquisition
The cat was immobilized as described in a previous section. Two different source-
detector positions were examined simultaneously where we placed the source fiber in
one position for example in b4 and then placed a detector fiber in position b2 and a
second in b6 following the grid system as described in the previous study. The
spectrometer was triggered at the beginning of each trial by an external pulse. Data
acquisition and analysis were performed by the Elantest software (This program is
available at ftp://www.lfd.uiuc.edu/lfd/egratton/elantest/), which communicated via a
68
USB II port with the HR4000 spectrometers where each spectrometer was operated
independently by its own laptop computer. Synchronization of the data acquisition with
the visual stimulation was achieved by an external Stanford Research (1290-D
Reamwood Avenue, Sunnyvale, CA 94089, USA) pulse generator (model DS345). This
was achieved by selecting the trigger mode of the function generator whereby the
external pulse that was provided by the start of the trial from the stimulus routine, was
then used as an input to the function generator where an arbitrary waveform was
generated to provide a continued positive TTL pulse that was split to deliver this
continuous trigger to each spectrometer for the duration of 15 seconds. This was critical
as this spectrometer required that the trigger be maintained for the length of time desired
for data acquisition. The trial lasted for 25 seconds so at the beginning of each trial this
trigger routine was reset. The trigger and synchronization system provided the correct
time signals to the Elantest software and the external trigger input port of the
spectrometer. The spectrometer acquired spectra every 5 ms. First, a reference spectrum
was taken under the condition of no trial, meaning that the cat observed the screen at its
normal background level as described in the previous section, but otherwise not
performing either task. Equal blocks of data consisting of 100 trials were collected under
different conditions. Data acquisition began with the beginning of each trial. The
collection of spectra was synchronized with an externally supplied pulse and lasted for
15seconds for both the VS and NVS trials. At the end of each trial, for each placement of
the source-detector, we have 100 X 20 = 2000 flashes for a 2 Hz stimulation (flashes)
frequency.
69
Figure 7.1- Schematic showing synchronization of visual stimulation and data acquisition
Figure 7.2- Diagram of experimental instrumentation
Source
Tungsten Lamp
DataAcquisition
DetectorHR4000
Spectrometer
DataAcquisition
DetectorHR4000
Spectrometer
Source
Tungsten Lamp
DataAcquisition
DetectorHR4000
Spectrometer
DataAcquisition
DetectorHR4000
Spectrometer
Start acquisition pulse
3500 pulses
0 15 seconds
10
Reward
15 second
Inter-trial
Visual Stimulation
500ms
Start acquisition pulse
3500 pulses
0 15 seconds
10
Reward
15 second
Inter-trial
Visual Stimulation
500ms
3000 Spectra
70
VII. E Data Analysis
In this chapter, we present the analysis of the fast hemodynamic signal. Data were
collected for the first 15seconds of each trial covering the full time of stimulation for VS
trials as well as 5 seconds where the reward (food) is given to the cat corresponding to a
total of 3000 spectra. Folding was done as a function of number of trials where the entire
trial (15 seconds) was block averaged to see if there were any consistent differences in
the spectral signals observed due to physiological changes occurring during the 100 trials.
A definitive spectral pattern and intensity pattern changes were observed in the raw data
matrix. We examined this raw data to interpret if the signals observed had physiological
significance.
VII. F Results
In this section, we focus on the raw data matrices observed by block averaging the entire
15seconds for each 100 trials both under VS and NVS conditions. The data is displayed
as maps where the intensities vary from a light blue (minimum value) to red (maximum
value), in some cases it is deliberately set to a scale where the values lie outside the
chosen scale to enhance the contrast for small signals. We present the data after we
perform the detrending routine to show the heartbeat data more clearly. The histogram
plot of the heartbeats is then presented for each visual stimulation frequency.
1. Raw Data- Block Averaged for entire trial
Two cases were investigated where in one case the cat was rewarded after 10 seconds of
the visual task and another where no reward (food) is given to the cat. The graphs are
atypical for most fields, so to explain the representation of data, we show intensity plots
where the horizontal axis from left to right gives temporal information while the bottom
71
corner of the vertical axis is the shortest wavelength investigated, 650nm increasing
along that axis to the top corner which is the longest wavelength, 1100nm. In the
following figure, (figure 7.3), we explain the data representation. If we examine the plot
choosing wavelengths that correspond to the tissue chromophores of physiological
importance in the NIR such as 690nm for HHb, 830nm for O2Hb, 960nm for water, we
can then determine the changes in these parameters as a function of time. The following
plot shows that for the first two seconds after the onset of the stimulus there is a decrease
in the transmission (increase in absorption) of the HHb (690nm) with a quasi-
simultaneous increase in transmission ( decrease in absorption) of the O2Hb (830nm)
while the water band remains relatively constant. At 2.5 seconds, the optical BOLD effect
is seen where the transmission due to the O2Hb decreases and the signal due to the HHb
increases showing that there is an increase in the absorption of the O2Hb with a decrease
in the HHb. The optical BOLD effect is maintained until the reward is given at 10
seconds, then a sharp transition (~ms) is seen that has no spectral dependence at 11.5
seconds. Then the initial dip (reverse BOLD effect) is seen where the O2Hb signal shows
an increase in transmission while the HHB shows a decrease in transmission.
72
Figure 7.3- Intensity Map describing spectral and temporal changes and the physiological importance.
In figure 7.4, we clearly see that in the panels corresponding to the VS trials, (b4b2 and
b4b6 VS food and no food), for the shorter wavelengths (650-700nm), there is an initial
decrease for the first 4 seconds followed by an increase from the time of 6 seconds after
the onset of stimulation and the reverse is seen for the higher wavelengths. This is the
classic BOLD effect. This is not seen in the NVS trials, (b4b2 and b4b6 NVS food and no
food) where the intensity maps are relatively flat up to the time of 10 seconds. However,
if we consider a trial where the reward is given, (b4b2 and b4b6 food), an additional
feature is seen at roughly 1 second after the reward is given to the cat at time 10 seconds
from the start of the trial in both the VS and NVS trials. There is a large signal which is
seen as an increase in the intensity of the shorter wavelengths and a simultaneous
decrease in the higher wavelengths (opposite to the BOLD effect), (b4b2 and b4b6 food,
VS and NVS). This effect is not seen in the trials where the reward is not given. The
Reward
14,0
00
12,0
00
10,0
00
8,0
00
6,0
00
4,0
00
2,0
00
0
680nm 1100nm820nm 960nm
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (red)
Sharp transition in spectral components with decrease in HHb (blue) and increase in O2Hb (red)- Initial Dip
ms
Reward
14,0
00
12,0
00
10,0
00
8,0
00
6,0
00
4,0
00
2,0
00
0
680nm 1100nm820nm 960nm
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (red)
Sharp transition in spectral components with decrease in HHb (blue) and increase in O2Hb (red)- Initial Dip
ms
14,0
00
12,0
00
10,0
00
8,0
00
6,0
00
4,0
00
2,0
00
0
680nm 1100nm820nm 960nm
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (red)
Sharp transition in spectral components with decrease in HHb (blue) and increase in O2Hb (red)- Initial Dip
14,0
00
12,0
00
10,0
00
8,0
00
6,0
00
4,0
00
2,0
00
0
680nm 1100nm820nm 960nm
14,0
00
12,0
00
10,0
00
8,0
00
6,0
00
4,0
00
2,0
00
014,0
00
12,0
00
10,0
00
8,0
00
6,0
00
4,0
00
2,0
00
0
680nm 1100nm820nm 960nm680nm 1100nm820nm 960nm680nm 1100nm820nm 960nm
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (red)
Sharp transition in spectral components with decrease in HHb (blue) and increase in O2Hb (red)- Initial Dip
ms
73
signal is spatially dependent but the trend is the similar at each source-detector pair
examined.
Figure 7.4- Comparison of Intensity Maps for source detector configurations under two conditions of food or no food presented to the cat.
In figure 7.5, we show the raw data obtained for 100 trials block averaged. In the left
panel, for the average change in intensity, a clean sharp transition is seen at 200ms after
the reward is given at 10 seconds, where the rise time of this transition was less than
100ms as shown in the expanded region in the graph below. The intensity map as a
function of wavelengths shown in the right panel shows that the optical BOLD effect (as
described previously) is seen at 2 seconds after the onset of the stimulation. More
importantly, there is a sharp change that is seen at roughly 300 ms after the reward is
B4b2 VS/ no food
B4b2 NVS/ no food
B4b2 VS/food
B4b2 NVS/food
14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000
Reward
1100nm
680nm
14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000
B4b6 VS/ no food
B4b6 NVS/ no food
B4b6 NVS/food
B4b6 VS/food
ms
ms1100nm
680nm
Reward
B4b2 VS/ no food
B4b2 NVS/ no food
B4b2 VS/food
B4b2 NVS/food
14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000
Reward
1100nm
680nm
14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000
B4b6 VS/ no food
B4b6 NVS/ no food
B4b6 NVS/food
B4b6 VS/food
ms
ms1100nm
680nm
Reward
74
administered that has no specific spectral dependence as shown by the insert below the
map that is flat across wavelengths. This sharp transition is seen for the NVS trial, where
the reward is given followed by the initial dip. If the reward is not given, this sharp
transition is not observed.
Figure 7.5- Left panel shows Average change in intensity for b6c5 and the zoomed region to show rise time of sharp transition, right panel shows the corresponding intensity map with an insert to show flat spectral dependence. If we examine the raw data from one trial, we see that this sharp transition is seen
independent of the type of trial (VS and NVS) as shown for b4c5 in the following figure
7.6. The optical BOLD effect is also seen in the VS trials as this fiber configuration lies
in the visual cortex. During the NVS trials we observe a flat spectrum until the sharp
transition at the time of the reward. The rise time for this sharp transition was determined
to be less than 100ms.
Rise Time ~100msRise Time ~100ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
680nm 1100nm
75
Figure 7.6- Raw Data and Intensity Maps for a single trial for b4b6 VS trial (top) and NVS (bottom) to show sharp transition and zoomed to show rise time
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-red)
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-red)
Rise Time ~100msRise Time ~100ms
Rise Time ~100msRise Time ~100ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
No Optical Bold Effect- Flat Spectrum
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
No Optical Bold Effect- Flat Spectrum
76
If we then look at the averaged data for 100 trials, we see that the sharp transition that is
seen in the single trial from both the VS and NVS trials, the sharp transition is maintained
as it is sharp after 100 trials as shown in figure 7.7.
77
Figure 7.7- Raw Data and Intensity Maps for 100 trials for b4b6 VS trial (top) and NVS (bottom) to show sharp transition and zoomed to show rise time
Rise Time ~110msRise Time ~110ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-blue)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-blue)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect-
Flat Spectrum
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect-
Flat Spectrum
Sharp transition in spectral components
Rise Time ~150msRise Time ~150ms
78
In the following plots, figure 7.8-7.10, we show the intensity maps showing the changes
in transmission signals as a function of both spectral and frequency of VS stimulation for
different source-detector positions. Figures 8 and 9 refer to configurations, b4b6 and b4c5
that lie in the visual cortex. One can see that in both cases, the optical BOLD effect is
observed where the onset is at 2 seconds after the beginning of the trial for b4b6 and 2.5
seconds for b4c5. The sharp spectrally independent transition is seen at all frequencies
after the reward is given. The reverse BOLD effect (where the increase in transmission
due to the O2Hb with a decrease in the HHb) is also seen and the magnitude of the effect
varies as a function of frequency of stimulus.
79
Figure 7.8- Intensity Maps as a function of stimulation frequency for b4b6 100 VS trials showing optical BOLD effect and sharp transition after the reward is given.
B4b6 Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
1Hz Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
2Hz
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
2.5HReward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
4Hz
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Sharp transition in spectral components
5Hz
80
Figure 7.9- Intensity Maps as a function of stimulation frequency for b4c5 100 VS trials showing optical BOLD effect and sharp transition after the reward is given.
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-red)
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-red)
1Hz Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-blue)
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-blue)
Sharp transition in spectral components
2Hz
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-red)
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect-decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (yellow-red)
2.5Hz Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effectdecrease in transmission in O2Hb(blue) while increase in transmission of HHb (yellow)
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effectdecrease in transmission in O2Hb(blue) while increase in transmission of HHb (yellow)
4Hz
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect-decrease in transmission in O2Hb ( dark blue) while increase in transmission of HHb (yellow-blue)
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Sharp transition in spectral components
Optical Bold Effect-decrease in transmission in O2Hb ( dark blue) while increase in transmission of HHb (yellow-blue)
5Hz
B4C5
81
In figure 7.10, a3a5, which lies in the motor cortex, shows that there is no observed
optical BOLD effect for the VS trials independent of stimulus frequency but there is a
sharp spectrally independent transition that is seen at roughly 500ms after the onset of the
reward.
82
Figure 7.10- Intensity Maps as a function of stimulation frequency for a3a1 100 VS trials showing no optical BOLD effect and sharp transition after the reward is given.
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
1Hz
A3A5
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
2Hz
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
2.5HzReward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
4Hz
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
No Optical Bold Effect
Sharp transition in spectral components
5Hz
83
In figures 7.11-7.13, we show the average change in intensities and we show that there is
a sharp transition for each configuration. The zoomed regions show that the rise time is
dependent on the spatial location is independent of the frequency of stimulus presented to
the cat.
84
Figure 7.11- Raw Data matrices showing zoomed inserts of rise time of sharp transitions for b4b6 100 VS trials as a function of stimulation frequency.
Rise time ~80msRise time ~80ms
Rise time ~80msRise time ~80ms
Rise time ~80msRise time ~80ms
Rise time ~80msRise time ~80msRise time ~80ms
Rise time ~80msRise time ~80ms
B4b6
85
Figure 7.12 - Raw Data matrices showing zoomed inserts of rise time of sharp transitions for b4c5 100 VS trials as a function of stimulation frequency.
Rise Time ~150msRise Time ~150ms
Rise Time ~110msRise Time ~110msRise Time ~110ms
Rise Time ~50msRise Time ~50msRise Time ~50ms
Rise Time ~110msRise Time ~110ms
Rise Time ~150msRise Time ~150ms
B4c5
86
Figure 7.13- Raw Data matrices showing zoomed inserts of rise time of sharp transitions for a3a1 100 VS trials as a function of stimulation frequency.
Rise time ~100msRise time ~100ms Rise time
~100msRise time ~100ms
Rise time ~100msRise time ~100msRise time ~100ms
Rise time ~100msRise time ~100ms
Rise time ~100msRise time ~100ms
A3a1
87
In the following curves, we show the comparison of the average intensities as a function
of spatial location in the cortex for two different stimulus frequencies 2 and 4Hz trials.
See figure 7.14. We see that the direction of the sharp transition as well as the time at
which it occurs is dependent on the position.
Figure 7.14- Comparison of Intensities vs time as a function of location for 2 and 4 Hz stimulation frequencies. Zoomed areas emphasize differences in latency time of the sharp transition In our setup, we are able to examine two source detector configurations simultaneously.
Graphs of synchronized locations are shown in figures 7.15. The graphs corresponding to
the motor cortex (a3a1 and a3a5) show that, while the general shape is the same, the
Comparison of Intensity as a function of location-2Hz
00.20.40.60.8
1
0 5000 10000 15000
Time(ms)
Nor
mal
ized
In
tens
ity
b4b22HzNorm b4b62HzNorm b4c52HzNormb4c62HzNorm a3a52HzNorm a3a12HzNorm
Intensity vs Time as a function of location-4Hz
00.10.20.30.40.50.60.70.80.9
1
0 2000 4000 6000 8000 10000 12000 14000
Time(ms)
Nor
mal
ized
Inte
nsity
a3a54HzNor a3a14HzNorm b4c5Norm b4c64HzNormc5b44hzNorm c5c64HzNorm b4b24hzNorm b4b64HzNormb6c54HzNorm b6b44HzNorm
Zoomed Comparison- 2Hz
0
0.2
0.4
0.6
0.8
1
9900 10100 10300 10500 10700 10900
Time(ms)
Nor
mal
ized
Inte
nsity
b4b22HzNorm b4b62HzNorm b4c52HzNormb4c62HzNorm a3a52HzNorm a3a12HzNorm
Zoomed Graph-4Hz
00.10.20.30.40.50.60.70.80.9
1
9900 10100 10300 10500 10700 10900
Time(ms0
Nor
mal
ized
Inte
nsity
a3a54HzNor a3a14HzNorm b4c5Norm b4c64HzNormc5b44hzNorm c5c64HzNorm b4b24hzNorm b4b64HzNormb6c54HzNorm b6b44HzNorm
Zoomed to show Transition-2 Hz
0
0.2
0.4
0.6
0.8
1
10000 10100 10200 10300 10400 10500
Time(ms)
Nor
mal
ized
Inte
nsity
b4b22HzNorm b4b62HzNorm b4c52HzNormb4c62HzNorm a3a52HzNorm a3a12HzNorm
Zoomed to show transition- 4Hz
00.10.20.30.40.50.60.70.80.9
1
10100 10150 10200 10250 10300 10350 10400 10450 10500
Time(ms)
Nor
mal
ized
Inte
nsity
a3a54HzNorm a3a14HzNorm b4c5Norm b4c64HzNormc5b44hzNorm c5c64HzNorm b4b24hzNorm b4b64HzNormb6c54HzNorm b6b44HzNorm
88
transition time differed by 300 ms where the a3a1 led the a3a5. However b4b6 and b4b2
have completely different trends where the transitions have different signs. For b4b2, the
signal is seen to increase while the signal for b4b6 decreases. Secondly, b4b2 shows a
significant peak at 10,400 ms that is not seen in b4b46.
Figure 7.15- Comparison of data sets acquired simultaneously from different locations, a3a1/a3a5 (motor cortex) and b4b2/b4b6 (visual cortex) to show differences in temporal behavior and spectral shape.
Comparison of a3a1 and a3a5Simultaneous Measurements
00.2
0.40.6
0.81
0 5000 10000 15000
Time(ms)
Norm
aliz
ed
Inte
nsiti
es
a3a52HzNorm
a3a12HzNorm
Comparison of b4b2 and b4b6Simultaneous Measurements
00.2
0.40.6
0.81
0 5000 10000 15000
Time(ms)
Norm
aliz
ed
Inte
nsiti
es
b4b22HzNorm
b4b62HzNorm
Zoomed- a3a1 and a3a5
0
0.2
0.4
0.6
0.8
1
9900 10400 10900
Time(ms)No
rmal
ized
Inte
nsiti
es
a3a54HzNor
a3a14HzNorm
Zoomed b4b2 and b4b6
0
0.2
0.4
0.6
0.8
1
9900 10400 10900
Time (ms)
Norm
aliz
ed In
tens
ities
b4b24hzNorm
b4b64HzNorm
Zoomed to show transtion- a3a1 and a3a5
0
0.2
0.4
0.6
0.8
1
10100 10200 10300 10400 10500
Time(ms)
Norm
aliz
ed In
tens
ity
a3a54HzNorma3a14HzNorm
Zoomed to show transition- b4b2 and b4b6
0.20.30.40.50.60.70.80.9
10100 10200 10300 10400 10500
Time(ms)
Norm
aliz
ed In
tens
ity
b4b24hzNorm
b4b64HzNorm
89
VII. G Discussion
There is a colossal signal that is seen roughly 0.5 seconds after the time that the reward is
given. The rise time of the signal varies from as little as 50 ms to 150ms dependent on the
source-detector location. This signal is also seen in the visual and frontal cortices and is
independent of the type of trial (VS and NVS) and of the frequency of stimulus presented
as shown in figures 7.12-7.15. This signal is absent when the food is not presented. The
transition is wavelength independent which is typical of a change due mainly to changes
in scattering, reminiscent of the fast signal as described by optical methods.61,62 This
proved that our current instrumental setup as well as broadband spectral technique is able
to detect fast changes and spectrally resolve them. More importantly, this sharp transition
is seen in as little as one trial. This is unheard in the field as for all methods to detect
neuronal activity reliably, employ thousands of trials are required to see this change.63-65
The question is why this signal is seen in all areas related to both motor and visual
cortices after the onset of the reward? Control experiments where the food was not
administered showed that this signal disappeared as shown in figure 7.4. First let us
examine the physiology, to explain why this effect is seen roughly 0.5 seconds after the
reward is given. At this time, the cat is undergoing several processes associated with
activation of the somatosensory cortex and motor cortex especially in the cases where
visual stimulation is presented, there is an additional response when the stimulation has
stopped that would result in an additional stimulus. At the time of the reward, a feeding
tube is presented to the cat touching its mouth initiating the feeding process. This tube is
presented with the same delay after the main task is completed (10 seconds of flashes if
VS trial). This precise synchronization could be at the origin of the sharp transition
90
observed after 100 trials. If we examine the adjacent cortices to the placement of our
tubes in both the visual and motor cortices, we can then determine if this signal is a direct
result of a physiological process. If we look at the schematic showing the placement of
the tubes with respect to the atlas of the cats’ brain, we see that the “a” row lies above the
border of the areas 4 and 6, which are motor cortices, posterior to this is the area 3 which
is the somatosensory cortex.66-67 Hence, any activation in these areas due to the reward
will be detected using our setup. The area of detection can be roughly described by a
circle of radius 1cm when our source-detector separation is 4mm. This explains the signal
that is seen in the “a” row but not in the visual cortices. However, upon further
examination of the areas adjacent to the visual cortex, activation in areas 5 and 7 will be
detected by our optical fibers. Areas 5 and 7 are considered the equivalent of parietal
cortex in the primate. Areas (17-19) are considered similar to occipital, and lateral
suprasylvian cortex similar to visual temporal cortex. They are sometimes referred to as
parietal association areas. They receive sensory input from visual, somatosensory, and
auditory lower-order areas, and project to sensory-motor cortex, particularly to areas 4
(mostly from area 5) and 6 (mostly from area 7).68-70 Hence, the signals that are seen are
indeed due to a physiological process originating in the somatosensory cortex and they
manifest as a fast signal due to scattering. This effect was not seen in the earlier study as
the length of data acquisition was restricted to the time of visual stimulation. Quite
frankly, it seems that the cat is more interested in the food than the visual stimulation, or
said a different way; a larger part of the brain is involved in the reward (just my luck).
Secondly, there are several signals that are superimposed on each other in the raw data;
however they occur at different time scales. There is the slow Optical BOLD effect that
91
was discussed in an earlier chapter, which is on the order of seconds, the pulse which has
a frequency of 3Hz (on the order of 300ms) that will be discussed in the next chapter and
a fast hemodynamic signal that is on the order of 100ms which is referred to as the initial
dip. Let us address each one in turn.
In the first study, we collected data for only 8 seconds of the trial and did not have access
to the reward period. The raw data again shows the optical BOLD effect as discussed
previously for VS trials in the visual cortex. If we look at figure 7.4, which are intensity
graphs of b4b2 VS and NVS, we can see that for the longer wavelengths that correspond
to the O2Hb, there is a decrease in the transmission as shown by the intense blue color,
which indicates an increase in absorption at these wavelengths roughly 2 seconds after
the onset of stimulation, with a quasi-simultaneous increase in transmission ( decrease in
absorption) as shown by the red color in the shorter wavelengths that correspond to the
HHb, the classic BOLD effect. The control experiments again show that the spectral
shape is flat for the NVS trials and the frontal lobes for all trials, VS and NVS. Further
examination of these graphs show that at the end of the visual task, the sharp transition is
seen at the time the reward is administered (at 10 seconds). Immediately following this
response ( on the order of 100ms), we can see that the reverse is true of the signals in that
there is an increase in the absorption due to HHb and a decrease in O2Hb, the initial dip.
This response lasts for approximately 1 second but the onset is less than 500 ms after the
sharp transition is seen. This is a proposed physiological response where there is a
consumption effect that is seen before the vasodilation occurs delivering the much needed
O2Hb. This is a precursor to the BOLD effect that will arise as a result of the
92
somatosensory activation. However, in the trials where the reward was administered and
where the fast signal is strongest, we see that the increase in HHb is larger.
VII. H Chapter Summary
This cat has yielded some novel and exciting results that has not been discussed
previously in the literature. We observed a “fast” signal (on the order of ms) in all areas
due to activity in the somatosensory cortex detected by spectral methods where the signal
is seen in a single trial. We observe that the signal is dependent on the location. The rise
time of the transition was seen to be from 50-150ms and the latency was also location
dependent. Our signal is highly reproducible yet it is seen in as little as one trial. The
optical BOLD effect is seen for the VS trials in the visual cortex and absent in the NVS
trials for the visual cortex and for all trials in the motor cortex confirming the results of
the earlier study. In summary, we can then follow a logical timeline of the physiological
processes: there is the slow optical BOLD effect that is on the order of seconds seen in
the VS trials for configurations in the visual cortex, followed by a sharp transition with a
rise time of less than 100ms seen in all configurations regardless of the trial type and
independent of stimulation frequency followed by the “initial dip” or reverse BOLD
effect which occurs within 500ms of the sharp transition. The control experiments again
show that the spectral shape is flat for the NVS trials and the frontal lobes for all trials,
VS and NVS.
93
VIII. Investigation of Cat Visual Cortex-Pulse
VIII. A Introduction
In the previous chapter, we described the novel finding of the fast signal that was
determined to be due to activation of the somatosensory cortex. Careful examination of
the raw data matrices showed that there were three signals of different physiological
origins and temporal distributions. We also discussed the slower hemodynamic response
that is the optical BOLD effect as well as a faster hemodynamic response that is the
initial dip. The last signal that has not been discussed is the heartbeat which is
hemodynamic but fast in that it occurs about every 300 milliseconds. In this chapter we
determine if the magnitude of this signal is comparable to that of the anticipated signal
due to the scattering changes following visual stimulation. The coupling of the pulse with
the visual stimulation and reward of the cat revealed that additional measures must be
taken to extract the fast neuronal signal due to visual stimulation. In this chapter, we
present simulations that show the SNR limitations of our equipment and of the true
physiology of the brain. We also present the analysis across flashes for both VS and NVS
trials both pre and post- pulse correction.
VIII. B Simulations
In our first study, we concluded that the SNR achieved using the original instrumentation
was insufficient to clearly see the fast neuronal signal. Here, we present simulations to
determine the limits of resolution for our current instrumental setup. If we consider a
sinusoidal signal as our test signal, the following discussion describes the logic. In the
following panels, 8.1, we simulate our fast neuronal signal which has a frequency of 2Hz
which is expected following a visual stimulation frequency of 2 Hz to the animal as
94
shown in the left panel. The right panel shows the result of our folding routine clearly
showing the recovery of the waveform.
Figure 8.1-Simulated Data with a sine wave of frequency 2 Hz as the fast signal due to visual stimulation shown on the left and the recovery of the waveform using our folding technique. If we simulate realistic data with the absolute worse SNR possible where the random
noise is 10% of the DC signal and the simulated fast signal is 0.0001 of the DC signal, we
see that we can still recover the underlying sinusoidal waveform. The Simulated noisy
data is shown in the left panel of figure 8.2 while the right panel shows the results of our
folding manipulation for 10 trials. Hence, it can be inferred that the signal with this SNR
can be clearly recovered after the folding of 100 trials.
Figure 8.2- Simulated Data of a sine wave of frequency 2 Hz where the Random Noise was 10% of DC signal on the left, right panel shows the recovery of the waveform after folding of 10 trials.
95
Now, we consider two signals that have similar frequencies. Physiologically these signals
correspond to the fast signal as described earlier and a hemodynamic signal due to the
pulse. We can separate the two signals with our folding routine provided that they are
equal in magnitude and there is no noise as shown in figure 8.3.
Figure 8.3- Simulated Data of two sine waves of equal magnitude superimposed one with frequency of 2 Hz to simulate fast signal and the second with frequency 3 Hz to simulate the pulse in top left panel. Top right panel shows the recovery of the 2Hz wave using the folding technique. Bottom center shows the recovery of 3 Hz using the folding technique. However, if the signals are not equal where the fast signal is less than 0.05 of the
magnitude of the signal due to the pulse, our folding routine cannot resolve the two
signals even in the absence of noise. This result is seen in the noise- free case and is
independent of the SNR as shown in figure 8.4.
96
Figure 8.4- Simulated Data of two sine waves of unequal magnitude superimposed one with frequency of 2 Hz to simulate fast signal and the second with frequency 3 Hz to simulate the pulse in left panel where the magnitude of the fast signal is 0.05 of the pulse signal. Right panel shows that we cannot recover the 2Hz wave using the folding technique. Hence, from our simulations we conclude that the folding routine and our instrumentation
has the necessary SNR to recover the fast signal due to visual stimulation with a horrible
SNR where the Noise is 10 % of the DC signal even if the magnitude of the fast signal is
0.0001 of the DC signal with the folding of 100 trials. However, with the addition of the
signal due to the pulse, a new complication is revealed in that the relative signal of the
pulse to the fast signal determines the resolution if they have similar frequencies (2Hz for
the fast signal and 3Hz for the pulse of a cat). Our simulations show that if the fast signal
is less than 0.05 of the pulse signal, we cannot resolve the two signals, independently, of
the SNR.
VIII. C Experimental Procedure
Methods- Cat Protocol, Visual Stimulation and Behavioral Paradigm and Technical
Aspects
The same cat was used for these experiments hence the preparation of the animal as
described previously is applicable for this study.
97
VIII. D Data Analysis
As stated previously, we block average 100 trials. We also performed a Fast Fourier
Transform (FFT) on this raw data to determine the frequency of a dominant signal. It
yielded a frequency of approximately 3 Hz that we identify with the heartbeat of the cat.
Additionally, after performing the folding manipulation of 100 trials, it was evident that
there was an underlying signal that followed the frequency of the heartbeat of the cat in
conjunction with the familiar optical BOLD effect discussed before. The observation of
the pulse after averaging one hundred independent trials was surprising and indicated that
there was some sort of synchronization of the pulse with the stimulation during the trial.
We reasoned that if the pulse (which is a relatively large signal) becomes synchronized
with the flashes, we will observe a signal that could be erroneously interpreted as due to
the neuronal response. The removal of the pulse is routinely done in experiments in
humans to avoid this possible artifact.71
Description of the Pulse Removal Routine
To better visualize the pulse in the optical signal, in the presence of the slow
hemodynamic response, we first performed an operation where we implemented a high-
pass filter to remove the slower signal (on the order of seconds). This process is referred
to as detrending in our notation. The raw data, after detrending and pulse removal showed
a very large pulse response. See figure 8.5. This routine was not simple because the heart
rate varied in frequency and amplitude and that the heartbeat itself was seen to have a
definite spectral shape. To remove the pulse, for each pulse we must identify the time at
which each pulse occurs during the trial. We achieve this by filtering the data sequence
using a 5th order Bessel band-pass filter in the range 2.4-4.3 Hz. This data is applied in
98
both directions to the data set to avoid introducing timing errors via a phase shift. The
time of the pulse was determined using the maxima and minima of the resulting filtered
signal. This normalized data (with respect to time) was then block averaged to obtain the
average shape of the pulse. The average shape of the pulse can be described by a
triangular wave which agrees with our expectations. The average wave was then adapted
to each pulse by interpolating the average wave shape to the actual duration of the pulse.
The adaptation routine performs a best fit of the amplitude of the pulse to the data during
each pulse. The amplitude of the fitted data as well as the timing of each pulse is then
recorded for further data analysis. Once, the heartbeat was extracted, we studied the
statistics of the pulse by constructing a histogram that recorded the time of heartbeats that
were seen during each trial for the full set of trials (100) and their amplitudes. However,
after performing the simulations, it was discovered that the stimulation frequencies
(except for the 5Hz) that were used in the trial were too close to the frequency of the
pulse.
A second approach was implemented to resolve the fast neuronal signal due to visual
stimulation. We employed Principle Component Analysis to determine the principle basis
components to determine if one of the major components followed the frequency of the
stimulation and if another was following the heartbeat. An external program,
Caterpillar.EXE was used for this analysis. We reasoned that this program will be able to
separate the fast signal for the visual stimulation frequencies that differed greatly from
the pulse frequency as the PCA routine shows only the orthogonal components.
99
VIII.E Results
Raw Data- Block Averaged – Detrended
After data detrending ( high- pass filtering), in figure 8.5, one can clearly see a periodic
signal shown by the intermittent areas of yellow that has a period of 3 Hz across the
entire trial independently of the presence of Visual stimulation. The signal increases by a
factor of 100 approximately 500 ms after the delivery of the reward (at 10s). This signal
is due to the somatosensory cortex and has been discussed in a previous chapter. This
pattern is seen in all of the source detector configurations both in the visual cortex and the
frontal lobes. It has a definite spectral shape and is still seen after block averaging of 100
trials hence it is synchronized with respect to the time of the trial.
Figure 8.5- Panels show detrended folded data with the pulses highlighted in both VS and NVS trials for different locations in the brain, motor and visual cortices. Scale set to emphasize differences in contrast between pulses and fast signal due to somatosensory after the reward is given at 10 seconds. Removal of the pulse and Analysis
We then applied the pulse removal routine to the first 10 seconds of the raw data. The
results are shown in figure 8.6 to show the effect of the pulse routine removal for b4b6
for a stimulation frequency of 4Hz for 100 VS trials.
14,00012,00010,0008,0006,0004,0002,0000
Scale -50-100
b4b6 1st10files
b4b6 last10files
b4b2 1st100files
b4b2 last100files
Scale -300-600
b4b6 1st100files
b4b6 last100files
14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000
Scale -50-100
b4b6 1st10files
b4b6 last10files
b4b2 1st100files
b4b2 last100files
Scale -300-600
b4b6 1st100files
b4b6 last100files
14,00012,00010,0008,0006,0004,0002,0000
b4c5 1st100files
b4c5 last100files
Scale -300-600
A3a5 1st100files
A3a5 last100files
14,00012,00010,0008,0006,0004,0002,0000 14,00012,00010,0008,0006,0004,0002,0000
b4c5 1st100files
b4c5 last100files
Scale -300-600
A3a5 1st100files
A3a5 last100files
Pulses
100
Figure 8.6- Left panel shows raw data for b4b6 VS trials for the 1st ten seconds (before reward) while the right panel shows the data post pulse-correction for a stimulation frequency of 4 Hz.
The following graphs in figure 8.7 show a histogram displaying the timing of the pulses
occurring during the entire trial. We reasoned that if the pulse will occur randomly (with
respect to the beginning of the trial), after plotting the time of the pulse in a histogram
where each time bin ( horizontal axis) is 1/8 of the period of the pulse, the histogram
should be relatively flat. Instead if the pulse occurs roughly at the same time with respect
to the beginning of the trial, then the histogram should show a definite pattern
corresponding to the synchronization of the pulse. Additionally, the graphs show eight
wavelength regions, (the entire spectrum was divided into eight equal regions) to show
the spectral variation as a function of time. The spike at 7.5 seconds is due to an artifact
that derives from the fact that we explore this time interval twice in our algorithm. The
artifact appears at the center of the time interval and it is due to double accounting of the
pulse in this region.
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Optical Bold Effect- decrease in transmission in O2Hb ( blue) while increase in transmission of HHb (blue-yellow)
b4b6VS
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
b4b6VS
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
Reward
14,00012,00010,0008,0006,0004,0002,0000
680nm
1100nm
820nm
960nm
ms
101
Figure 8.7- Left panel top shows the histogram of the timing of the pulses for visual cortex (b4b6) 100 VS trials while the Bottom panel shows the corresponding histogram for the NVS trials. Right panel-Top show the histogram for motor cortex (a3a1) 100 VS trials while the Bottom panel shows the corresponding histogram for the NVS trials. All histograms obtained using the pulse correction routine for a 4 Hz stimulation frequency. In figure 8.7, it is clear that the synchronization of the pulses is highest at the start of the
trial from 0.5 seconds after the onset of stimulation for about 2 seconds for b4b6VS and
NVS trials (visual cortex) and a3a1VS and NVS (motor cortex) trials and at a second
period of roughly 1 second before the reward is given at the 10 second mark.
102
Results- Fast Signal due to Visual Stimulation
Removal of the pulse for the analysis of the Fast Signal by Folding
Here, we present folded data pre and post pulse correction. The graphs are of the average
intensity as a function of folded period for all flashes corresponding to a total of 10
seconds. In the following graphs, we show that for a frequency stimulus of 2.5 Hz
corresponding to a period of 400 ms (200ms on, 200 ms off), we can see for the source-
detector configurations that lie in the visual cortex (b4b6, b4c5), we can see that there is a
peak that has a width of approximately 100ms for the VS trials while the profile is flat for
the NVS trials, however for the frontal lobes, a3a5, the profile is flat independent of type
of trial. The graphs show the comparison between each data set. However, the change is
only on the order of a few parts per ten thousand. Figures 8.8 and 8.9
Figure 8.8- Top panels show the fast signal analysis using the folding routine across a period of 400ms corresponding to 2.5 Hz for both VS and NVS trials, right side compares coronal and parasagital regions of visual cortex ( b4c5 and b4b6), while right panel compares motor cortex with visual cortex ( b4c5 and a3a5).
103
If we compare pre and post pulse corrected data, in the following figures shows the
effect of the routine. The following graphs show that the pulse correction period
changes the data by a minimal amount.
Figure 8.9-Graphs comparing the folding results pre and post pulse correction routines, Top panel shows the pre- pulse correction routine for visual and motor cortex for both VS and NVS trials. Bottom panel shows the results of folding after pulse correction routine was applied to the data.
104
Analysis of the fast signal by PCA
Here we present data that was processed using the Principle Component Analysis (PCA)
program. The graphs show the component that showed the visual stimulation for the trials
considered. The following figure 8.10 shows the PCA done for the visual and motor
cortex for a stimulation frequency of 5Hz. It shows that there is a periodic structure only
in the case of the VS trial for the visual cortex while the other components have no
structure comparable to the stimulation frequency for both the VS and NVS trials for the
motor cortex (a3a1) and the NVS trial for the visual cortex (b4b6)
Figure 8.10- Results of the PCA program applied to b4b6 shown in the left panel and a3a1 in the right for both VS (top) and NVS (bottom) trials for a stimulation frequency of 5 Hz. Black line indicates the period corresponding to a folding time of 200ms. Double the period is shown to determine if the periodic structure repeats itself.
105
VIII. F Discussion
The heartbeat was clearly seen as we have very high spatial resolution as we probe short
source-detector distances. This signal proved to be significant and is coupled to other
physiological processes. The problem becomes how we can decouple the heartbeat from
the signals as a function of several wavelengths if the heartbeat itself has a spectral shape.
The pulse is seen in all trials and in all cortex areas examined however the magnitude is
cortex dependent. The frequency of this pulse and the synchronization of the pulse
presented possible complications in the detection of the fast neuronal signal during the
VS. Statistical analysis of the pulse as shown in histograms in figures 8.7 show that the
pulse is synchronized at the beginning of the trials for VS trials in the visual cortex
(b4b6) while it is random for the NVS trials. However, in the motor cortex, (a3a1), this
synchronization is seen in both VS and NVS trials. This synchronization is independent
of the frequency of the stimulus presented to the animal. Secondly, the pulse correction
routine appears to have little effect on the data as shown in figure 8.9. At this time, we
believe that this synchronization is an artifact introduced by the pulse correction routine
and that we must fully understand the effect of the pulse correction routine that is
currently used. Clearly modifications must be made. Furthermore, the question arises if
this pulse signal is larger than the proposed fast neuronal signal that is purely a scattering
signal and how do we fully remove this signal to see the signal that has been reported in
the literature to be due to a change in scattering. We implemented a pulse correction
routine on the raw but detrended data as shown in figure 8.7, the data post correction
show that the pulse was indeed removed from the tasks. Simulations showed that the
magnitude of the pulse compared to that of the fast signal is the deciding factor in being
106
able to separate the two signals if they are close in frequency for example 2Hz for the fast
signal and 3Hz for the pulse. The SNR of the system has no impact on this conclusion.
We concluded that new data manipulations must be used to extract the scattering signal.
We applied the PCA program to the 5Hz as the PCA is able to separate components if
they are orthogonal with respect to each other. Also, with this program we can determine
if one component follows the timing of the pulse and a second one following the timing
of the fast signal. We reasoned that the 5 Hz stimulation frequency was sufficiently larger
than the frequency of the pulse to ensure orthogonal components. Hence, only for this
stimulation frequency we observed the fast signal following visual stimulation as shown
in figure 8.10. Now, the question becomes why is this signal due to the somatosensory
cortex is seen in as little as one trial and not the signal due to the visual cortex during the
VS trials. However, it is known that the fast signal in the somatosensory cortex is larger
than that due to the visual cortex.72
Future studies must then take this signal to pulse ratio for data analysis techniques.
Secondly, the frequency of stimulation presented to the subject must be at least a factor of
1.3 the frequency of the pulse of the subject in this case would correspond to that of the
cat.
VIII. G Chapter Summary
In conclusion, the simulated data showed that we have the required SNR to recover the
fast signal if the signal is at least 0.0001 of the DC signal. A novel complication was
revealed that the criteria that determined the ability to resolve the fast signal in the
presence of the signal of the pulse was in fact the ratio of magnitudes of these two
signals. It was determined that if the fast signal was below 0.05 of the pulse signal that
107
folding was insufficient to resolve these two signals if they were perfect sinusoidal waves
of fixed frequencies. A second parameter of importance was the ratio of the frequencies
of the pulse to the frequency of the fast signal which follows the stimulation frequency
presented to the animal. The coupling of these frequencies was an additional
complication where it was determined that this ratio must be at least 1.3. For this ratio,
we performed PCA on this data as we surmised that the signals should be orthogonal for
the stimulation frequency of 5 Hz. We observed a component that follows the
stimulation frequency only for the visual cortex for VS trials while no component was
observed in the motor cortex VS and NVS trials and the NVS trials for the visual cortex.
Additionally, more work must be done to debug the pulse correction routine and
determine the true magnitude of the fast signal as compared to that of the pulse.
108
IX. Summary
In the introduction, I made a claim that the Physicists provide significant contributions to
the field of Biology and Medical Sciences by applying basic physics principles to the
field. Specifically, in this work, we probed the light-matter interactions in the NIR region
to understand physiological processes in the mammalian brain. We sought to improve on
existing principles and propose a new technique by which we can decipher these
processes spectrally. This technique touted to be independent of the light transport regime
allowed us to examine the hemodynamics and neuronal activity.
A new technique was developed to provide complementary information to the existing
Multi- Distance Frequency Domain Photon Migration technique that has been the bread
and butter of the Gratton Laboratory. The group has been the pioneers of the photon
migration field, so it is only natural to continue the tradition of completely
revolutionizing the field. The new technique involves using a broad-band spectrum to
probe the tissue and then recovering the light that has traversed the tissue using an ultra-
fast (5ms/spectra) spectrometer. We examine the raw data by deconvolution into the
individual basis spectra associated with the tissue chromophores, O2Hb, HHb, Water, fat.
We account for scattering by a mathematical description of λ-n , where the n coefficient is
allowed to vary to account for different types of scattering that is present in the probed
volume. Hence, scattering is separated from absorption by its distinctive spectral shape as
opposed to the “time of flight” delay that was previously used.
The aim was then to test this technique and see if it produces results that were
comparable to the well established Fd- NIRS in distinguishing physiological processes.
Secondly, we wanted to prove that this technique was light transport regime independent
109
which is not the case for the Fd-NIRS. The cat was chosen as an ideal test subject as its
anatomy is such that photons are not fully diffusive before being detected as the total size
of the grey matter in the cat is roughly 3mm thick. Additionally, we had a priori
information about the activation of the visual cortex as a response to specific stimuli.
First approach was to build a phantom to simulate the anatomy of the smaller brain by
simulating the optical properties accounting for the heterogeneities. The results of that
experiment were important for the following reasons: one to determine if we had
sufficient SNR and to determine the effects that the heterogeneous nature of the brain in
the different light transport regimes (non- diffusive- diffusive). It was determined that
the true nature of tissue can have a significant effect on the observed optical signals;
hence we must understand the type and size of the effect of the different types of tissue
under investigation. It is clear that for the probing of the brain, not only is the source
detector separation important to determine depth of tissue penetration but the position of
the white matter with respect to the grey matter. Additionally, we have sufficient SNR for
the animal study.
Our experiment on the animal was novel in that it was the first study on an awake animal
where a broad-band spectral approach is used to determine the individual NIR spectrum
of tissue components. It was able to detect physiological changes by spectral methods,
reminiscent of the fMRI BOLD signal where we refer to our observation as the optical
BOLD effect. Our technique was seen to be independent of the light modality (diffusive
or non-diffusive), as we were able to recover accurate changes in the brain. For the
application in a mammalian brain, we have examined the behavior of the scattering,
O2Hb and HHB (BOLD effect) simultaneously with other tissue components such as
110
water content. Additionally, there was a large change in the water signal that could not be
accounted physiologically. The behavior of water during stimulation has not been
discussed in previous literature. We proposed a model to show that this signal was due to
the heterogeneous nature of the brain and the effects that physiological processes like
vasodilation can have on the observed optical signals. Our model separates the brain into
two compartments optically opaque and transparent. We proposed that the water change
can be an indication on the amount that the vessels dilate upon stimulation as we know
that optically the large blood vessels appear opaque so any changes as seen in
vasodilation will result in a decrease in all chromophores. Furthermore, the technique has
proven to have high enough temporal and spatial resolution to adequately determine the
localized hemodynamics.
As is standard in our field, one must validate any model using phantoms. Our claim that
the water signal is due to the vasodilation raised some concerns in that vasodilation or
any process that involves the change of sizes of the opaque tissue (large blood vessels) to
the transparent tissue (capillaries) can skew the true origin of the signals. The question
arises, is this effect comparable to the real signals and is it possible to distinguish this
effect fro a “true” signal? We built a phantom to simulate vasodilation under the
conditions of different sizes of vessels modeling them as optically opaque. Our results
from the dynamic phantom made of the various diameters of opaque spokes show that
“vasodilation” per se, as simulated by changing the diameter of the spokes, could cause
an apparent decrease in all spectral components, including water and scattering.
Therefore in the presence of vasodilation of the large blood vessels, all spectral
components should decrease proportionally to their contribution to the overall spectrum.
111
We propose that the apparent decrease in water content (and partially scattering)
observed in the cat brain following visual stimulation is in fact due to vasodilation. In
this model, the time course of the water change reflects the dynamical changes of the
diameter of the blood vessels. We also determined that the relative changes due to
vasodilation should be about 4% for superficial blood vessels but less than 0.4% for
vessels at 6 mm or more from the surface. In the case of the cat brain, the changes due to
vasodilation could be more significant in the measurements due to the small size of the
brain with respect to humans. Note, that this artifact in the estimation of spectral
components due to vasodilation would not have been recognized if we had only used a
few wavelengths. In fact, using a broad band spectral analysis allowed us to distinguish
changes that equally affect all spectral components from specific changes affecting only
one spectral component. Vasodilation decreases the total light transmission and reduces
the spectral amplitude. Therefore, if we measure only relative spectral changes, we will
measure a reduction of the spectral amplitude at all wavelengths. As a consequence of
our studies, we conclude that determination of chromophore concentration in tissue using
only few wavelengths (for O2Hb, HHb and scattering) is not sufficient to characterize the
origin of the changes.
We went one step further in our phantom studies to validate our technique to show that
we an accurately recover the changes in scattering and absorption separately. In our
phantom, we mimic spectral changes in absorption and scattering by inserting spokes
which contain different concentrations of absorbers and scatterers. We show that we can
accurately recover (in two distinct cases) the spectrum (which is measured independently
using a spectrophotometer) of the absorber and the spectral component due to scattering
112
as a function of concentration (of absorber and scatterer). The range of concentrations
was restricted to the regime of low optical density. It was observed that in the diluted
absorber/small scattering changes regime, the relative changes and concentration of
absorbers/scatterer as recovered by the measurement method followed a linear
relationship. When the concentration of the absorbers and scattering centers increases,
the optical changes were no longer proportional to the concentration of the
absorber/scatterer and the changes start to saturate the measured absorption. Our results
can have implications on the way one interprets changes in chromophore components in
the presence of vasodilation, or any other physiological condition that changes the
relative contribution of the “opaque” compartment.
In our final quest to validate our technique as a viable one to determine the physiological
processes associated with brain activation, we pursued the detection of the fast neuronal
signal that is believed to be related to a change in scattering. We revisit the cat to
continue our analysis. The first study was limited by the SNR of our instrumentation,
while we were able to detect the slower optical BOLD signal, we could not conclude
definitively that we observed the fast signal. New instrumentation with higher sensitivity
was obtained as well as the protocol changed to significantly improve our detection
sensitivity. We also chose to collect data for a longer period to have additional access to
other physiological processes. We also adopted the collection of data for two different
locations using two separate but synchronized spectrometers. We were able to obtain
results where this cat has yielded some novel and exciting results that has not been
discussed previously in the literature. We observed a “fast” signal (on the order of ms) in
all areas due to activity in the somatosensory cortex detected by spectral methods where
113
the signal is seen in a single trial. We observed that the signal is dependent on the
location. The rise time of the transition was seen to be from 50-150ms and the latency
was also location dependent. Our signal is highly reproducible yet it is seen in as little as
one trial. The optical BOLD effect is seen for the VS trials in the visual cortex and absent
in the NVS trials for the visual cortex and for all trials in the motor cortex confirming the
results of the earlier study. Careful examination of the raw data matrices showed that
there were three signals of different physiological origins and temporal distributions. In
summary, we were able to follow a logical timeline of the physiological processes: there
is the slow optical BOLD effect that is on the order of seconds seen in the VS trials for
configurations in the visual cortex, followed by a sharp transition with a rise time of less
than 100ms seen in all configurations regardless of the trial type and independent of
stimulation frequency followed by the “initial dip” or reverse BOLD effect which occurs
within 500ms of the sharp transition. The control experiments again show that the
spectral shape is flat for the NVS trials and the frontal lobes for all trials, VS and NVS.
The last signal that proved to be of importance was that due to the pulse. This pulse
signal was observed to be a large signal and observed in all cortices under all stimulation
frequencies independent of VS or NVS trials.
Consequently, we performed simulations to calculate the SNR required of our new
instrumentation in the determination of the fast signal. We concluded that the folding
routine and our instrumentation has the necessary SNR to recover the fast signal due to
visual stimulation with a horrible SNR where the Noise is 10 % of the DC signal even if
the magnitude of the fast signal is 0.0001 of the DC signal with the folding of 100 trials.
However, with the addition of the signal due to the pulse, a new complication is revealed
114
in that the relative signal of the pulse to the fast signal determines the resolution if they
have similar frequencies (2Hz for the fast signal and 3Hz for the pulse of a cat). Our
simulations show that if the fast signal is less than 0.05 of the pulse signal, we cannot
resolve the two signals, independently, of the SNR. Consequently, we concluded that new
data manipulations must be used to extract the scattering signal. We applied the PCA
method to the 5Hz as the PCA is able to separate components if they are orthogonal with
respect to each other. Also, with this program we can determine if one component
follows the timing of the pulse and a second one following the timing of the fast signal.
We applied a pulse correction routine to our data in an attempt to decouple the pulse from
the fast signal. This routine was complicated as the pulse itself had a distinctive spectral
shape. However, we were able to study the statistics of the extracted pulse and observed a
synchronization of the pulse at the beginning of the trials for VS trials in the visual cortex
while it is random for the NVS trials. However, in the motor cortex, this synchronization
is seen in both VS and NVS trials. This synchronization is independent of the frequency
of the stimulus presented to the animal. We believe that the observed synchronization
could be an artifact of the pulse correction routine. We reasoned that the 5 Hz stimulation
frequency was sufficiently larger than the frequency of the pulse to ensure orthogonal
components. Hence, only for this stimulation frequency we observed the fast signal
following visual stimulation. Now, the question becomes why is this signal due to the
somatosensory cortex is seen in as little as one trial and not the signal due to the visual
cortex during the VS trials. However, it is known that the fast signal in the somatosensory
cortex is larger than that due to the visual cortex.
115
In conclusion, the studies of this dissertation have provided important information on the
different physiological processes as well as provided basic understanding of the spectral
behavior associated with them. The novelties are several and in casual mention include
the detection of the fast signal in as little as one trial (unheard of!), detection of
physiological processes such as the BOLD effect and fast signal shown as changes in
scattering by a spectral method that is independent of the light transport regime, the
observation of the debated “initial dip” as well as the synchronization of the pulse by
spectral methods.
116
X. Future Plans
This chapter should really be referred to as if I had a time machine what would I have
done differently. One pressing issue has been the correct removal of the pulse and the
determination of the magnitude of the fast signal due to the visual stimulation as
compared to the pulse. In retrospect, the obvious solution would be to use an external
device to monitor the cat as is done in human studies. Secondly, the exploration of the
PCA as a viable form of data acquisition must be pursued to be more quantitative in our
analysis. This study was done on one cat, at some point it would be nice to examine other
animals to validate the findings within the species as well as apply the technique to other
species. Clearly much more work must be done to fully understand the physiological
processes and the appropriate forms of data analysis
117
References
1. Wyatt J.S., Cope M., Delpy D.T., Wray S. & Reynolds E.O. Quantification of
cerebral oxygenation and haemodynamics in sick newborn infants by near
infrared spectrophotometry. Lancet. 1986 Nov 8;2(8515):1063-6.
2. De Blasi R.A., Fantini S., Franceschini M.A., Ferrari M. & Gratton E. Cerebral
and muscle oxygen saturation measurement by frequency-domain near-infra-red
spectrometer. Med Biol Eng Comput. 1995 Mar;33(2):228-30.
3. Lewy WJ, Levin S, Chance B. Near -infrared measurement of cerebral
oxygenation. Correlation with electroencephalographic ischemia during
ventricular fibrillation. Anesthesiology 1995, 83 : 738-746
4. Gopinath SP, Robertson CS, Contant CF, Narayan RK, Grossman RG, Chance B.
Early detection of delayed traumatic intracranial hematomas using near-infrared
spectroscopy. J Neurosurg 1995; 83:438-44.
5. Alfano, R. R., Pradhan, A., Tang, G. C. & Wahl, S. J. Optical spectroscopic
diagnosis of cancer and normal breast tissues. J. Opt. Soc. Am. B 1989 vol. 6, page
1015.
6. Andersson-Engels S., Johansson J., Svanberg S. & Svanberg K. Fluorescence
diagnosis and photochemical treatment of diseased tissue using lasers: Part I.
Anal Chem. 1989 Dec 15;61(24):1367A-1373A
7. E. M. Sevick, B. Chance, J. Leigh, S. Nioka, and M. Maris, Quantitation of time-
and frequency-resolved optical spectra for the determination of tissue
oxygenation. 1991 Anal. Biochem., 195, 330-351,
118
8. Chance B, Wang NG, Maris M, Nioka S, Sevick E. Quantitation of tissue optical
characteristics and hemoglobin desaturation by time- and frequency-resolved
multi-wavelength spectrophotometry. Adv Exp Med Biol. 1992;317:297-304
9. http://omlc.ogi.edu/classroom/ece532/class3/scatterers.html
10. S. Chandrasekhar (1960). Radiative Transfer, Dover Publications, Inc
11. G Zaccanti, E Battistelli, P Bruscaglioni and Q Wei. Analytic relationships for the
statistical moments of scattering point coordinates for photon migration in a
scattering medium. 1994 Pure Appl. Opt. 3 897-905
12. A. J. Welch, M. J. C. van Gemert, W. Star, and B. C. Wilson, Definitions and
overview of tissue optics in Optical-Thermal Response of Laser-Irradiated Tissue
(Plenum, New York. 1995), pp. 3268-3277
13. Van der Zee, P., Essenpreis, M. & Delpy, D.T. 1993 Optical properties of Brain
Tissue. Proc. SPIE 1888, 454-465.
14. Ishimaru, A. 1978 Wave propagation and scattering in random media, vol. 1. New
York: Academic Press.
15. Patterson, M. S., Chance, B & Wilson, B. C. Time-resolved reflectance and
transmittance for the non-invasive measurement of tissue optical properties. 1989
Appl. Opt. 28, 231-236.
16. Fishkin, J. B. and Gratton, E. Propagation of photon density waves in strongly
scattering media containing an absorbing semi- infinite plane bounded by a
straight edge. 1993 J. Opt. Soc. Am. A 10, 127-140.
119
17. Gratton, E., Fantini, S., Franceschini, M. A., Gratton, G. & Fabiani, M.
Measurements of scattering and absorption changes in muscle and brain. 1997
Phil. Trans. R. Soc. Lond. B 352, 727-735.
18. Wolf, M., M. A. Franceschini, L. A. Paunescu, V. Toronov, A. Michalos, U.
Wolf, E. Gratton, and S. Fantini. Absolute frequency-domain pulse oxymetry of
the brain: Methodology and measurements. In: Proc. ISOTT Oxygen Transport to
Tissue XXII. 27th Annual Meeting of ISOTT, J. F. Dunn, H. M. Swartz, eds.
(2000).
19. K. Tanner, E. D’Amico, A. Kaczmarowski, S. Kukreti, J. Malpeli, W.W.
Mantulin, E. Gratton. “Spectrally resolved neurophotonics: a case report of
hemodynamics and vascular components in the mammalian brain.” Journal of
Biomedical Optics, in press.
20. D. Purves et al, Neuroscience.
21. Gratton, E., Fantini, S., Franceschini, M. A., Gratton, G. & Fabiani, M.
Measurements of scattering and absorption changes in muscle and brain. 1997
Phil. Trans. R. Soc. Lond. B 352, 727-735.
22. E. Okada, M. Firbank, M. Schweiger, S. R. Arridge, M. Cope, and D. T. Delpy.
“Theoretical and experimental investigation of near-infrared light propagation in a
model of the adult head.” Appl. Opt. 36, 21-31 (1997).
23. H. Cui and J. Malpeli, “Activity in the Parabigeminal Nucleus during eye
movements directed at moving and stationary targets,” J. Neurophysiol. 89, 3128-
3142, (2003).
120
24. A. Tate and J. Malpeli, “Effects of Focal Inactivation of Dorsal or Ventral layers
of the Lateral Geniculate Nucleus on Cat’s ability to see and fixate small targets,”
J. Neurophysiol. 80, 2206-2209 (1998).
25. F. Reinoso- Suarez, Topographischer Hirnatlas der Katze, fur Experimental-
Physciologische Unstersuchungen. T. 4 E. Merck AG Darmstadt (1961).
26. A. Shmuel and A. Grinvald, “Functional Organization for Direction of Motion
and Its Relationship to Orientation Maps in Cat Area 18,” J. of Neurosci 16,
6945-6964 (1996).
27. V. Toronov, A. Webb, J. H. Choi, M. Wolf, L. Safonova, U. Wolf and E. Gratton,
“Study of Local Cerebral Hemodynamic Fluctuations by Simultaneous
Frequency-Domain near-infrared spectroscopy and fMRI,” Optics Express 9, 417-
427, (2001).
28. S. Ogawa, R.S. Menon, S. G. Kim and K. Ugurbil, “On the Characteristics of
Functional Magnetic Resonance Imaging of the brain.” Annu. Rev. Biophys.
Biomol. Struct. 27, 447-74, (1998).
29. L. Leybaert. “Neurobarrier coupling in the brain,” Journal of Cerebral Blood
Flow & Metabolism, 25, 2-16 (2005).
30. http://library.thinkquest.org/C0115080/?c=vessels
31. F. A. Marks. “Optical spectroscopy of breast biopsies and human breast cancer
and xenografts in nude mice.” Frontiers in Bioscience 2, 1-10 (1998).
121
32. A. Grinvald, E. Lieke, R.D. Frostig, C.D. Gilbert, and T. N. Wiesel, “Functional
architecture of cortex revealed by optical imaging of intrinsic signals,” Nature
324, 361-364 (1986).
33. Walker, S. A., S. Fantini, and E. Gratton. “The effect of index of refraction
mismatch on the recovery of optical properties of cylindrical inhomogeneities in
an infinite turbid medium.” SPIE Proc., Optical Tomography and Spectroscopy of
Tissue: Theory, Instrumentation, Model, and Human Studies II, 2979 219-225
(1997).
34. Franceschini, M. A., S. Fantini, A. Cerussi, B. Barbieri, B. Chance and E.
Gratton. “Quantitative spectroscopic determination of hemoglobin concentration
and saturation in turbid medium: Analysis of the effect of water absorption.”
Journal of Biomedical Optics 2, 147-153 (1997).
35. M. Firbank and D. T. Delpy. “A design for a stable and reproducible phantom for
use in near infra-red imaging and spectroscopy.” Phys. Med. Biol. 38, 847–853
(1993).
36. E. Okada, M. Firbank, M. Schweiger, S. R. Arridge, M. Cope, and D. T. Delpy.
“Theoretical and experimental investigation of near-infrared light propagation in a
model of the adult head.” Appl. Opt. 36, 21-31 (1997).
37. J. R. Mourant, I. J. Bigio, D. A. Jack, T. M. Johnson, and H. D. Miller,
“Measuring absorption coefficients in small volumes of highly scattering media:
source-detector separations for which path lengths do not depend on scattering
properties.” Appl. Opt. 36, 5655–5661 (1997).
122
38. V. Venugopalan, J. S. You, and B. J. Tromberg, “Radiative transport in the
diffusion approximation: an extension for highly absorbing media and small
source–detector separations.” Phys. Rev. E 58, 2395–2407 (1998).
39. R. Nossal, J. Kiefer, G. H. Weiss, R. Bonner, H. Taitelbaum, and S. Havlin,
“Photon migration in layered media,” Appl. Opt. 27, 3382–3391 (1988).
40. H. Taitelbaum, S. Havlin, and G. H. Weiss, “Approximate theory of photon
migration in a two-layer medium.” Appl. Opt. 28, 2245–2249 (1989).
41. E. Okada, M. Firbank, M. Schweiger, S. R. Arridge, J. C. Hebden, M. Hiraoka,
and D. T. Delpy, “Experimental measurements on phantoms and Monte Carlo
simulation to evaluate the effect of inhomogeneity on optical pathlength,” in
Optical Tomography, Photon Migration, and Spectroscopy of Tissue and Model
Media: Theory, Human Studies, and Instrumentation, B. Chance and R. R.
Alfano, eds., Proc. SPIE 2389,174–181 (1995).
42. S. T. Flock, S. L. Jacques, B. C. Wilson, W. M. Star, and M. J. C. van Gemert.
“The optical properties of Intralipid: a phantom medium for light propagation
studies.” Lasers Surg. Med. 12(5), 510-9 (1992).
43. V. Toronov, S. Walker, R. Gupta, J. H. Choi, E. Gratton, D. Hueber, A. Webb.
“The roles of changes in deoxyhemoglobin concentration and regional cerebral
blood volume in the fMRI BOLD signal.” NeuroImage 19, 1521–1531 (2003).
44. J. H. Choi, “Cerebral hemodynamics and oxygenation: Study of Baseline
spontaneous oscillation.” PhD Dissertation Thesis, Biophysics and Computational
Biology Program, University of Illinois at Urbana-Champaign, 2004.
123
45. A. Grinvald, E. Lieke, R.D. Frostig, C.D. Gilbert, and T. N. Wiesel, “Functional
architecture of cortex revealed by optical imaging of intrinsic signals,” Nature
324, 361-364 (1986).
46. N. V. Swindale, D. Shoham, A. Grinvald, T. Bonhoeffer and M. Hübener, “Visual
cortex maps are optimized for uniform coverage,” Nature Neuroscience 3, 2000,
822-826 (2000).
47. D. Jancke, F. Chavane, S. Naaman and A. Grinvald, “Imaging cortical correlates
of illusion in early visual cortex,” Nature 428, 423-426 (2004).
48. J. Frahm, H. Bruhn, M.L. Gyngell ,K. D. Merboldt , W. Hanicke and R. Sauter,
“Localized proton NMR spectroscopy in different regions of the human brain in
vivo. Relaxation times and concentrations of cerebral metabolites,” Magn Reson
Med. 1989 Jul; 11(1):47-63.
49. P. B. Barker , B. J. Soher , S. J. Blackband , J. C. Chatham , V. P. Mathews and
R. N. Bryan, “Quantitation of proton NMR spectra of the human brain using
tissue water as an internal concentration reference,” NMR Biomed. 1993 Jan-
Feb;6(1):89-94.
50. D. E.J. Linden, D. Prvulovic, E. Formisano, M. Völlinger, F. E. Zanella, R.
Goebel and T. Dierks, “The Functional Neuroanatomy of Target Detection: An
fMRI Study of Visual and Auditory Oddball Tasks,” Cerebral Cortex Volume 9,
8 : 815-823
51. Gratton, G., Goodman-Wood, M. R., and Fabiani, M. (2001), “Comparison of
Neuronal and Hemodynamic Measure of the Brain Response to Visual
Stimulation: an Optical Imaging Study,” Human Brain Mapping, 13, pp.13-25.
124
52. G. Gratton, M. Fabiani, P. M. Corballis, D. C. Hood, M. R. Goodman-Wood, J.
Hirsch, K. Kim, D. Friedman and E. Gratton, “Fast and Localized Event-Related
Optical Signals (EROS) in the Human Occipital Cortex: Comparisons with the
Visual Evoked Potential and fMRI,” NeuroImage 6, 168–180 (1997)
53. A. Villringer and B. Chance, “Non-invasive optical spectroscopy and imaging of
human brain function,” TINS Vol. 20, No. 10, 1997
54. M. A. Franceschini, V. Toronov, M. Filiaci, E. Gratton, and S. Fantini, “On-line
optical imaging of the human brain with 160-ms temporal resolution,” Optics
Express Vol. 6, No. 3, 2000
55. L. B. Cohen and R. D. Keynes, “Changes in light scattering associated with the
action potential in crab nerves,” J. Physiol. Lond 212, 259–275, 1971.
56. L. B. Cohen, R. D. Keynes, and L. Landowne, “Changes in light scattering that
accompany the action potential in squid giant axons: potential-dependent
components,” J. Physiol. Lond 224, 701–725, 1972.
57. D. M. Rector, R. F. Rogers, J. S. Schwaber, R. M. Harper and J. S. George,
“Scattered-Light Imaging in Vivo Tracks Fast and Slow Processes of
Neurophysiological Activation,” NeuroImage 14, 977–994, 2001
58. K.M. Carter, J.S. George and D.M. Rector, “Simultaneous birefringence and
scattered light measurements reveal anatomical features in isolated crustacean
nerve,” Journal of Neuroscience Methods 135, 9–16 , 2004
59. R.B. Buxton, “The Elusive Initial Dip,” NeuroImage 13, 953-958, 2001.
125
60. Vanzetta I and Grinvald A, “Evidence and lack of evidence for the initial dip in
the anesthetized rat: implications for human functional brain imaging,”
Neuroimage 13, 959-67, 2001
61. G. Gratton, M. Fabiani, D. Friedman, M. A. Franceschini, S. Fantini, P. M.
Corballis, and E. Gratton, “Rapid Changes of Optical Parameters in the Human
Brain During a Tapping Task,” J. Cognitive Neuroscience 7,446-456 (1995).
62. M.A. Franceschini and D.A. Boas, “Noninvasive measurement of neuronal
activity with near-infrared optical imaging,” Neuroimage, 2004
63. M. Wolf, U. Wolf, Jee H. Choi, V. Toronov, L. Adelina Paunescu, A. Michalos,
and E. Gratton, “Fast cerebral functional signal in the 100-ms range detected in
the visual cortex by frequency-domain near-infrared spectrophotometry,”
Psychophysiology Vol. 40, 521, 2003
64. Tasaki, I, “Rapid structural changes in nerve fibers and cells associated with their
excitation processes,” Japanese Journal of Physiology, 49, 125–138, 1999.
65. Gratton, G., and Fabiani, M. “Shedding light on brain function: The event-related
optical signal,” Trends in Cognitive Sciences, 5, 357–363, 2001.
66. A. Nieoullon and L. Rispal-Padel, “Somatotopic localization in cat motor cortex”,
Brain Research Vol 105, 405-422, 1976
67. M. Glickstein, “How are visual areas of the brain connected to motor areas for the
sensory guidance of movement?” Trends Neurosci. 23, 613–617, 2000
68. S. G. Lomber, D. Bertram and R. Payne, “Cerebral Areas Mediating Visual
Redirection of Gaze: Cooling Deactivation of 15 Loci in the Cat,” Journ of Comp.
Neurology 474:190–208, 2004
126
69. M. Steriade, A. Nunez and F. Amzica, “A novel slow (< 1 Hz) oscillation of
neocortical neurons in vivo: depolarizing and hyperpolarizing components,”
Journal of Neuroscience, Vol 13, 3252-3265, 1993
70. A.K. Engel, P.R. Roelfsema, P. Fries, M. Brecht and W. Singer, “Role of the
temporal domain for response selection and perceptual binding,” Cerebral Cortex,
Vol 7, 571-582, 1997
71. G. Gratton and P. M Corballis, “Removing the heart from the brain:
Compensation for the pulse artifact in the photon migration signal,”
Psychophysiology 32, 292–299, 1995
72. J. Steinbrink, M. Kohl, H. Obrig, G. Curio, F. Syre, F. Thomas, H. Wabnitz, H.
Rinneberg and A. Villringer, “Somatosensory evoked fast optical intensity
changes detected non-invasively in the adult human head,” Neuroscience Letters
291 105-108, 2000.
127
Author’s biography
Kandice Tanner was born on 14th April, 1980 in Port of Spain, Trinidad. She attended
high school in Trinidad where she focused on languages and science for her studies at
the advanced level for the Cambridge Examinations. She attained full distinctions and
secured a presidential scholarship to study at South Carolina State University,
Orangeburg, SC in 1998. She was a presidential scholar for four years. At South
Carolina State, she received the highest academic awards for outstanding engineering
student and top academic honors as a student athlete. She graduated in May, 2002,
summa cum laude with a dual degree in Electrical Engineering Technology and
Physics. She then attended University of Illinois- Urbana-Champaign to pursue her
graduate studies in Physics in August, 2002. She attained her Masters degree in
August, 2003 at the same time she switched from her focus of condensed Matter
physics to Biophysics under the tutelage of Dr. Enrico Gratton. She worked on Brain
Imaging until she attained her doctoral degree in July, 2006. She has accepted a
position as a Post-Doctoral researcher at the University of California, Irvine.