1.

Absorption
1.1. Some biological chromophores
1.2. Definition and units of absorption coefficient a [cm-1]
1.3. Example: bilirubin
1.4. Absorption spectra for biological tissues
2. Scattering
2.1. Some biological scatterers
2.2. Definition and units of scattering coefficient s [cm-1]
2.3. Definition of anisotropy g [dimensionless]
2.3.1. Scattering functions
2.3.2. Isotropic scattering function
2.3.3. Henyey-Greenstein scattering function
2.4. Reduced scattering coefficient s ' = s (1 - g) [cm-1]
2.5. Mie theory model for scattering
2.5.1. Brief overview of Mie theory math
2.5.2. Example calculation of angular scattering pattern
2.5.3. Example calculation of anisotropy
2.5.4. Example calculation of scattering cross section and coefficient
2.5.5. Scattering versus wavelength for 3 particle sizes
2.6. Mie scattering from cellular structure of soft tissues
2.7. Mie scattering from collagen fibers of dermis

1 Absorption
A simple analogy for the absorption of light by molecules is placing two identical bells side by side.
When one bell is rung, the other will sympathetically ring at the same frequency due to transfer of
energy from the struck bell. The resonance of the second bell matches the frequency of the first
ringing bell and hence the second bell accepts energy from the struck bell.
Photons are electromagnetic waves with a particular frequency. Molecules are a system with charge
separation (negative electron field and positive nucleus). The state of the molecular charge separation
can change in a quantized fashion by "absorbing" the energy of a photon. The photon frequency must
match the "frequency" associated with the molecule's energy transition in order for energy transfer
to occur. The relation between frequency and energy is
Energy = h(frequency) = hc/(wavelength)
where Energy is in [J], photon frequency is in [cycles per s] or [s -1], photon wavelength is in [m], c
is the speed of light in vacuo (c = 3.0x108 [m/s]), and h is Planck's constant (h = 6.62618x10-34 [J s]).
Unlike the bell analogy, photon absorption occurs as a quantum event, an all or none
phenomenon. Example: absorption of 514 nm photon from an argon ion laser by a hemoglobin
molecule.
In biomedical optics, absorption of photons is a most important event:
Absorption is the primary event that allows a laser or other light source to cause a
potentially therapeutic (or damaging) effect on a tissue. Without absorption, there is no
energy transfer to the tissue and the tissue is left unaffected by the light.
Absorption of light provides a diagnostic role such as the spectroscopy of a tissue. Absorption
can provide a clue as to the chemical composition of a tissue, and serve as a mechanism of
optical contrast during imaging. Absorption is used for both spectroscopic and imaging
applications.

Example: Heme absorption of photon

For example, a 514 nm photon from an argon ion laser when absorbed by a hemoglobin
molecule will transfer energy to the hemoglobin molecule:
hc/(0.514x10-1 [m]) = 3.86x10-19 [J]
Doesn't sound like much energy, does it? But consider the energy from the perspective of the
absorbing chromophore. The photon is absorbed by the heme chromophore within
the hemoglobin protein. The heme choromophore is roughly 1 nm is size.
After the heme chromophore absorbs one photon, the jump in energy density is roughly the
photon energy divided by a nm3 volume:
(3.86x10-19 [J])/(10-7 [cm])3 = 387 [J/cm3]

For comparison, the energy density of boiling water is 418 J/cm 3. So one photon carries quite
a lot of energy from the perspective of a small molecule!

1.1 Some biological chromophores

Molecules that absorb light are called chromophores. There are two major types of choromphores:
electronic transitions
vibrational transitions
Electronic transitions

There are many biological molecules which can absorb light via electronic transitions. Such
transitions are relatively energetic and hence are associated with absorption of ultraviolet, visible and
near-infrared wavelengths. The molecules generally have a string of double bonds whose pi-orbital
electrons act similar to the electrons in a metal in that they collectively behave as a small antenna
which can "receive" the electromagnetic wave of a passing photon. If the resonance of the pi-orbital
structure matches the photon's wavelength then photon absorption is possible.
In early biological evolution, the pyrrole molecule was a chromphore which could absorb sunlight
which enabled subsequent synthetic reactions that produced biological polymers and other protometabolic products. Combining four pyrroles into a tetrapyrrole ring (porphyrin) yielded an efficient
chromophore for collecting solar photons. Chlorophyll is such a porphyrin. Hemoglobin, vitamin
B12, cytochrome C, and P450 are also examples of porphyrins in biology. The figure lists some
common biological chromophores and shows some of their structures. Also see the
website spectra which is a compilation of chromophores, most absorbing in the ultraviolet and
visible.

Vibrational transitions

The field of infrared spectroscopy studies the variety of bonds which can resonantly vibrate or twist
in response to infrared wavelengths and thereby absorb such photons. Perhaps the most dominant
chromophore in biology which absorbs via vibrational transitions is water. In the infrared, the
absorption of water is the strongest contributor to tissue absorption.

Some vibrational frequencies

bond

cycles/cm,

wavelength,

= 1/

C-H stretch

2850-2960 [cm-1] 3.378-3.509 [m]

C-H bend

1340-1465

6.826-7.462

C-C stretch,bend 700-1250

8.000-14.29

C=C stretch

1620-1680

5.952-6.173

C=C stretch

2100-2260

4.425-4.762

CO32-

1410-1450

6.897-7.092

NO3

1350-1420

7.042-7.407

NO2

1230-1250

8.000-8.130

SO4 2-

1080-1130

8.850-9.259

O-H stretch

3590-3650

2.740-2.786

C=O stretch

1640-1780

5.618-6.098

N-H

3200-3500

2.857-3.125

ref: PW Atkins, "Physical Chemistry," p. 576, W.H. Freeman and Co., 1978.

1.2 Definition and units of absorption coefficient a [cm-1]

Consider a chromophore idealized as a sphere with a particular geometrical size. Consider that this
sphere blocks incident light and casts a shadow, which constitutes absorption. This description is of
course an incorrect and schematicized version of the real situation. However, it does provide a simple
concept which captures the essence of the absorption coefficient, the parameter we use to describe
the effectiveness of absorption.
The size of the absorption shadow is called the effective cross-section ( a [cm2]) and can be smaller
or larger than the geometrical size of the chromophore (A [cm 2]), related by the proportionality
constant called the absorption efficiency Qa [dimensionless]:

The absorption coefficient a [cm-1] describes a medium containing many chromophores at a

concentration described as a volume density a [cm3]. The absorption coefficient is essentially the
cross-sectional area per unit volume of medium.

Experimentally, the units [cm-1] for a are inverse length, such that the product aL is dimensionless,
where L [cm] is a photon's pathlength of travel through the medium. The probability of survival (or
transmission T) of the photon after a pathlength L is:
This expression for survival holds true regardless of whether the photon path is a straight line or a
highly tortuous path due to multiple scattering in an optically turbid medium.

1.3 Example: Bilirubin

The bilirubin molecule is a chromophore encountered when a newborn infant suffers from "jaundice",
a syndrome in which the skin presents a yellow color. Bilirubin is a breakdown product of
hemoglobin. Often there is significant hemolysis of red blood cells during child birth contributing to
a transient bilirubin load. Normally, bilirubin binds to the serum protein albumin and is carried to the
liver where enzymes convert it into a water-soluble form which is removed from the blood into the

bile. But in newborns who have not yet developed sufficient enzymes to accomplish this task, the
bilirubin accumulates in the blood, exceeds the holding capacity of the albumin, and spills into the
skin to give the yellow skin color, and into the brain to cause irreversible brain damage (kernicterus).
See article on optically monitoring bilirubinemia.
Let's approximate the values of A, a, a, Qa, and a for bilirubin.

The structure of bilirubin is shown. The diameter is

approximately 1 nm. So the geometrical area is A =
4.5x10-15 cm2 .

At 460 nm, the extinction coefficient of bilirubin is =

53,846
[cm-1 M-1 ]
(see
bilirubin
spectrum).
The extinction coefficient
[cm-1 M-1 ] quoted in the
literature is based on spectrometer measureme nts
reported as T = 10- CL where C is concentration [M] and
L is pathlength [cm]. Therefore,
a = C ln(10)

A typical jaundiced neonate might have a bilir ub in

concentration of 10 mg/dL, or (0.100 g/liter)/(574.65
g/mole) = 0.17x10-3 M. In such a case, the bilir ub in
absorption coefficient at 460 nm is roughly
a = C ln(10) = (53846 [cm-1 M-1 ])(0.17x10-3 M)(2.3)
= 21 cm-1 .
The
concentration
C
is
equivalent
to
a=
(0.17x10-3
[moles/liter])(6x1023
3
cm /liter] = 1.02x1017 [cm-3 ]

[mole-1 ])/(1000

The efficiency of absorption is estimated:

Qa = a/( aA) = (21 [cm-1 ])/((1.02x1017 [cm-3 ])(A =
4.5x10-15 [cm2 ])) = 0.046
The effective cross-section is a = Q aA = (0.046)(4.5x1015 [cm2 ]) = 2.1x10-16 [cm2 ]. Bilirubin's effective crosssectional diameter is sqrt(.046) or 21% the size of its
geometrical diameter.

Keep in mind that the wavelength of blue light (460 nm) is about 460-fold greater than the diameter
of the bilirubin chromophore. So collection of the electromagnetic wave of a photon by the "antenna"
of bilirubin is analogous to a how small radio antenna collects the electromagnetic wave from a radio
station (a 1000 MHz radio frequency -> 300 m wavelength). The concept of a "shadow" cast by an
opaque chromophore is merely a memory device to remember the definitions.

1.4 Absorption spectra for biological tissues

The figure below shows the primary absorption spectra of biological tissues. Also shown are the
absorption coefficients at some typical laser wavelengths.

There are several major contributors to the absorption spectrum:

In the ultraviolet, the absorption increases with shorter wavelength due to protein, DNA and
other molecules.
In the infrared, the absorption increases with longer wavelengths due to tissue water content.
Scaling the pure water absorption by 75% mimics a typical tissue with 75% water content.
In the red to near-infrared (NIR), absorption is miminal. This region is called the diagnostic
and therapeutic window (originally by John Parrish and Rox Anderson).
Whole blood is a strong absorber in the red-NIR regime, but because the volume fraction of
blood is a few percent in tissues, the average absorption coefficient that affects light transport
is moderate. However, when photons strike a blood vessel they encounter the full strong
absorption of whole blood. Hence, local absorption properties govern light-tissue interactions,
and average absorption properties govern light transport.
Melanosomes are also strong absorbers. However, their volume fraction in the epidermis may
be quite low, perhaps several percent. So again, the local interaction of light with the
melanosomes is strong, but the melanosome contribution to the average absorption coefficient
may modestly affect light transport.

2 Scattering
Scattering of light occurs in media which contains fluctuations in the refractive index n, whether such
fluctuations are discrete particles or more continuous variations in n.
In biomedical optics, scattering of photons is an important event:
Scattering provides feedback during therapy. For example, during laser coagulation of tissues,
the onset of scattering is an observable endpoint that correlates with a desired therapeutic goal.
Scattering also strongly affects the dosimetry of light during therapeutic procedures that
depend on absorption. The scattering affects "where" the absorption will occur.
Scattering has diagnostic value. Scattering depends on the ultrastructure of a tissue, eg., the
density of lipid membranes in the cells, the size of nuclei, the presence of collagen fibers, the
status of hydration in the tissue, etc. Whether one measures the wavelength dependence of
scattering, the polarization dependence of scattering, the angular dependence of scattering, the
scattering of coherent light, scattering measurements are an important diagnostic tool.
Scattering is used for both spectroscopic and imaging applications.

2.1 Some biological scatterers

The light scattered by a tissue has interacted with the ultrastructure of the tissue. Tissue ultrastructure
extends from membranes to membrane aggregates to collagen fibers to nuclei to cells. Photons are
most strongly scattered by those structures whose size matches the photon wavelength.

Scattering of light by structures on the same size scale as the photon wavelength is described by Mie
theory. Scattering of light by structures much smaller than the photon wavelength is called the
Rayleigh limit of Mie scattering, or simply Rayleigh scattering. The figure designates the size range
of tissue ultrastructure which affects visible and infrared light by Mie and Rayleigh scattering.
Here are some examples of structures which scatter light (click on figure to expand):

Mitochondria
Mitochondria are intracellular organelles about 1 m in length
(variable) which are composed of many folded internal lipid
membranes called cristae, as shown in the electron micrograph at
left. The basic lipid bilayer membrane is about 9 nm in width. The
refractive index mismatch between lipid and the surrounding
aqueous medium causes strong scattering of light. Folding of lipid
membranes presents larger size lipid structures which affect longer
wavelengths of light. The density of lipid/water interfaces within the
mitochondria make them especially strong scatterers of light.
(Drawing and micrographs from A. L. Lehninger, "Biochemistry", Worth Publishers, 1970.)

Collagen fibers, fibrils, and fibril periodicity

Collagen fibers (about 2-3 m in diameter) are composed of
bundles of smaller collagen fibrils about 0.3 m in diameter
(variable), as shown in the electron micrograph at left. Mie
scattering from collagen fibers dominates scattering in the infrared
wavelength
range.
On the ultrastructural level, fibrils are composed of
entwined tropocollagen molecules. The fibrils present a banded
pattern of striations with 70-nm periodicity due to the staggered
alignment of the tropocollagen molecules which each have an
electron-dense head group that appears dark in the electron
micrograph. The periodic fluctuations in refractive index on this
ultrastructureal level appear to contribute a Rayleigh scattering
component that dominates the visible and ultraviolet wavelength
ranges.

2.2 Definition and units of scattering coefficient

s [cm-1]

Consider a scattering particle idealized as a sphere with a particular geometrical size. Consider that
this sphere redirects incident photons into new directions and so prevents the forward on-axis
transmission of photons, thereby casting a shadow. This process constitutes scattering. This
description is of course an oversimplified and schematicized version of the real situation. However,
it does provide a simple concept which captures the essence of the scattering coefficient, a parameter
analogous to the absorption coefficient discussed previously.
The size of the scattering shadow is called the effective cross-section ( s [cm2]) and can be smaller
or larger than the geometrical size of the scattering particle (A [cm 2]), related by the proportionality
constant called the scattering efficiency Qs [dimensionless]:

The scattering coefficient s [cm-1] describes a medium containing many scattering particles at a
concentration described as a volume density s [cm3]. The scattering coefficient is essentially the
cross-sectional area per unit volume of medium.

Experimentally, the units [cm-1] for s are inverse length, such that the product sL is dimensionless,
where L [cm] is a photon's pathlength of travel through the medium. The probability of transmission
T of the photon without redirection by scattering after a pathlength L is:

2.3 Definition of anisotropy g [dimensionless]

The anistoropy, g [dimensionless], is a measure of the amount of forward direction retained after a
single scattering event. Imagine that a photon is scattered by a particle so that its trajectory is
deflected by a deflection angle , as shown in the figure below. Then the component of the new
trajectory which is aligned in the forward direction is shown in red as cos( ). On average, there is an
average deflection angle and the mean value of cos( ) is defined as the anisotropy. 13

A scattering event causes a deflection at angle from the original forward trajectory. There is also
an azimuthal angle of scattering, .
But it is the deflection angle which affects the amount of forward direction, cos( ), retained by the
photon.
Consider an experiment in which a laser beam strikes a target such as a cylindrical cuvette containing
a dilute solution of scattering particles. The scattering pattern p( ) is measured by a detector that is
moved in a circle around the target while always facing the target. Hence the detector collects light
scattered at various deflection angles in a horizontal plane parallel to the table top on which the
apparatus sits. The proper definition of anisotropy is the expectation value for cos( ):

It is common to express the definition of anisotropy in an equivalent way:

2.3.1 Scattering functions

The angular dependence of scattering is called the scattering function, p( ) which has units of [sr-1]
and describes the probability of a photon scattering into a unit solid angle oriented at an
angle relative to the photons original trajectory. Note that the function depends on only on the
deflection angle and not on the azimuthal angle . Such azimuthally symmetric scattering is a
special case, but is usually adopted when discussing scattering. However, it is possible to consider
scattering which does not exhibit azimuthal symmetry. The p( ) has historically been also called the
scattering phase function.
The scattering can be described in two ways:
Plotting p( ) indicates how photons will scatter as a function of
in a single plane of
observation (source-scatterer-observer plane). This pattern is similar to the type of goniometric
scattering experiments commonly conducted.
Plotting p( )2 sin indicates how photons will scatter as a function of the deflection
angle regardless of the azimuthal angle , in other words integrating over all possible in an
azimuthal ring of width d and perimeter 2 sin at some given . The p( )2 sin goes to zero
at 0 because the azimuthal ring becomes vanishingly small at 0. This plot is related to the
total energy scattered at a given deflection angle and hence is more pertinent to the value of
anisotropy.

Figure
depicts
a
typical
forward-directed
scattering
pattern
p( )
corresponding to an experimental goniometric measurement in a single source-scatterer-observer
plane, and p( )2 sin which integrates over all possible azimuthal angles .

2.3.2 Isotropic scattering function

An isotropic scattering function would scatter light with equal efficiency into all possible
directions. Such a scattering function would have the form:

2.3.3 Henyey-Greenstein scattering function

enyey and Greenstein (1941) devised an expression which mimics the angular dependence of light
scattering by small particles, which they used to describe scattering of light by interstellar dust clouds.
The Henyey-Greenstein scattering function has proven to be useful in approximating the angular
scattering dependence of single scattering events in biological tissues.
The Henyey-Greenstein function allows the anisotropy factor g to specify p( ) such that calculation
of the expectation value for cos( ) returns exactly the same value g. In other words, Henyey and
Greenstein devised a useful identity function. The Henyey-Greenstein function is:

It is common practice to express the Henyey-Greenstein function as the function p(cos );

A series of Henyey-Greenstein functions are shown in the following figure. The forward direction
along the original photon trajectory is 0. Scattering in the backward direction is 180. The curve for
g = 0 has a constant value of 1/4 .

2.4 Reduced scattering coefficient

s' = s(1 - g) [cm-1]
The reduced scattering coefficient is a lumped property incorporating the scattering coefficient
s and the anisotropy g:
s' = s(1 - g) [cm-1]

The purpose of s' is to describe the diffusion of photons in a random walk of step size of 1/s' [cm]
where each step involves isotropic scattering. Such a description is equivalent to description of
photon movement using many small steps 1/s that each involve only a partial deflection angle if
there are many scattering events before an absorption event, i.e., a << s'. This situation of
scattering-dominated light transport is called the diffusion regime and s'is useful in the diffusion
regime which is commonly encountered when treating how visible and near-infrared light propagates
through biological tissues.
The following figure shows the equivalence of taking 10 smaller steps of "mean free path" mfp =
1/s with anisotropic deflection angles and one big step with a "reduced mean free path" mfp' = 1/s'.

The following figure shows how many such big steps involving isotropic scattering are equivalent to
many small anisotropic steps:

2.5 Mie theory model for tissue optical properties

Mie theory describes the scattering of light by particles. "Particles" here means an aggregation of
material that constitutes a region with refractive index (np ) that differs from the refractive index of
its surroundings (nmed). The dipole reradiation pattern from oscillating electrons in the molecules of
such particles superimpose to yield a strong net source of scattered radiation. Also, the reradiation
patterns from all the dipoles do not cancel in all but the forward direction of the incident light as is
true for homogneous medium, but rather interfere both constructively and destructively in a radiation
pattern. Hence, particles "scatter" light in various directions with varying efficiency.
Gustav Mie in 1908 published a solution to the problem of light scattering by homogeneous spherical
particles of any size. Mie's classical solution is described in terms of two parameters, nr and x:

the magnitude of refractive index mismatch between particle and medium expressed as the
ratio of the n for particle and medium,
nr = np /nmed

the size of the surface of refractive index mismatch which is the "antenna" for reradiation of
electromagnetic energy, expressed as a size parameter (x) which is the ratio of the meridional
circumference of the sphere (2 a, where radius = a) to the wavelength ( /nmed) of light in the
medium,
x = 2 a/( /nmed)
A Mie theory calculation will yield the efficiency of scattering which relates the cross-sectional area
of scattering, s [cm2], to the true geometrical cross-sectional area of the particle, A = a2 [cm2]:
s = Qs A
Finally, the scattering coefficient is related to the product of scatterer number density, s [cm-3], and
the cross-sectional area of scattering, s [cm2], (see definition of scattering coefficient):
s = s s
Before using Mie theory to approximate the scattering behavior of biological tissues, let's
briefly examine the Mie calculation

illustrate the behavior of Mie scattering with some example calculations and figures

2.5.1 The math of Mie scattering

Consider a source, a spherical scattering particle, and an observer whose three positions define a
plane called the scattering plane. Incident light and scattered light can be reduced to their
components which are parallel or perpendicular to the scattering plane. As shown in the following
figure, the parallel and perpendicular components can be experimentally selected by a linear
polarization filter oriented parallel or perpendicular to the scattering plane.

The Scattering matrix describes the relationship between incident and scattered electric field
components perpendicular and parallel to the scattering plane as observed in the "far-field"
(ref: Bohren and Huffman):

The above expression simplifies in practical experiments:

The exponential term, -exp(-ik(r-z))/ikr, is a transport factor that depends on the distance
between scatterer and observer. If one measures scattered light at a contant distance r from the
scatterer, eg., as a function of angle or orientation of polarization, then the transport factor
becomes a constant.
The total field (Etot ) depends on the incident field (E i), the scattered field (Es) , and the
interaction of these fields (Eint ). If one observes the scattering from a position which avoids
Ei, then both Ei and Eint are zero and only Es is observed.
For "far-field" observation of Es at a distance r from a particle of diameter d such that kr >>
nc2, k = 2 / , nc = d/ , the scattering elements S3 and S4 equal zero (see Eq. 4.75, Bohren and
Huffman).
Practical experiments measure intensity, I = <E E*> = (1/2)a2, where E = a exp(-i ), and a is
amplitude and is phase of the electric field.
Hence for practical scattering measurements, the above equation simplifies to the following:

Mie theory yields two sets of descriptors of scattering:

ANGULAR SCATTERING PATTERN OF POLARIZED LIGHT Mie theory calculates the
angular dependence of the two elements, S1( ) and S2( ), of the Scattering matrix, from which the
scattered intensities of polarized light are computed (see example). The scattering pattern is also used
to calculate the anisotropy, g, of scattering by the particle.
Example of angular scattering calculation
Example of anistoropy calculation

EFFICIENCIES OF SCATTERING AND ABSORPTION Mie theory calculates the efficiencies

Qs and Qa of scattering and absorption, respectively, such that a = QaA and s = QsA, where A is
the geometrical cross-sectional area a2 for a sphere of radius a.
Example of Qs and s calculation

2.5.2 Angular patterns of Mie scattering

Consider the scattering pattern from a 0.304-m-dia. nonabsorbing polystyrene sphere in water
irradiated by HeNe laser beam:
np = 1.5721

particle refractive index

nmed = 1.3316

medium refractive index

a = 0.152 m,

particle radius

= 0.6328 m

wavelength in vacuo

As calculated before:
nr = np/nmed = 1.5721/1.3316 = 1.1806

relative refractive index

x = 2 a/( /nmed) = (2)(3.1415)(0.152)/(0.6328/1.3316) = 2.0097 size parameter

Run the Mie theory algorithm:

Mie(nr,
x)
--->
Mie(1.1806,

2.0097)

--->

S1( ),
S1.re

S1( )
jS1.im,

S2.re

as
+

complex
jS2.im

as

numbers

functions

of

To view the results, calculate the intensities of scattering for parallel and perpendicular
orientations of polarized source/detector pairs:
Ipa r
=
S2S2*
=
Re{(S2.re
+
jS2.im)(S2.re
jS2.im)}
Iper = S1S1* = Re{(S1.re + jS1.i m)(S1.re - jS1.im)}
which are shown in the following figures, and can be experimentally measured as described
below.
The following figures describe the experimental measurements that illustrate the angular dependence
of the scattered intensities Ipar and Iper:
Iper
Irradiate dilute solution of spheres in water with laser beam
polarization oriented perpendicular to the table. Collect
scattered light as a function of angle in plane parallel to
table. Place linear polarization filter in front of
detector perpendicular to the table.

Ipar
Irradiate dilute solution of spheres in water with laser beam
polarization oriented parallel to the table. Collect scattered
light as a function of angle in plane parallel to table. Place
linear polarization filter in front of detector parallel to the
table.

Example results: Polar and xy plots of scattering pattern for I par and Iper

Polar

plot
I( )

plot

Click figure to enlarge

Click figure to enlarge

For a randomly polarized light source, the total scattered light intensity is given by the term S 11:
S11 is the first element of the so-called Mueller Matrix, a 4x4 matrix which relates an input vector of
Stokes parameters (Ii, Qi, Ui, Vi) describing a complex light source and the output vector (Is, Qs, Us,
Vs) describing the nature of the transmitted light. For randomly polarized light, S 11 describes the
transport of total intensity:
Is = S11Ii

2.5.3 Mie theory calculation of anisotropy (g)

Let's use the scattering pattern calculated previously for a 0.304-m-dia. nonabsorbing polystyrene
sphere in water irradiated by HeNe laser beam to calculate the anisotropy (g) of scattering.
The definition of anisotropy guides the calculation of g using the function S 11( ) which describes the
scattered intensity for randomly polarized light:

The denominator in the above equation assures proper normalization of p( ):

When numerically evaluating the above expression for g, a large number of angles need to be
calculated, typically about 200 angles, in order to achieve a value of g with precision to at least 4
significant digits. For our example 0.304-m sphere, the calculation of g based on S11( ) yields g =
0.6608.
Note that the definition of g as cited in these notes assumes azimuthal symmetry, hence we have
calculated the g for randomly polarized light. Due to the symmetry presented by a spherical particle,
the g refers to scattering into all azimuthal angles regardless of the linear polarization of the incident
light.

2.5.4 Example: Mie calculation of Q s and s

Consider the scattering of a HeNe laser beam by 0.304-m-dia. nonabsorbing polystyrene spheres in
water at a concentration of 0.1% volume fraction:
np = 1.5721

particle refractive index

nmed = 1.3316 medium refractive index

a = 0.152 m, particle radius
= 0.6328 m wavelength in vacuo
fv = 0.001

volume fraction of particles in medium

Calculate the following parameters:

refractive index mismatch ratio, size parameter, geometrical cross-sectional area, and number density
nr = np /nmed = 1.5721/1.3316 = 1.1806

x = 2 a/( /nmed) = (2)(3.1415)(0.152)/(0.6328/1.3316) = 2.0097

A = a2 = (3.1415)(0.152)2 = 0.0726 m2
3
3
-3

s = fv/((4/3) a ) = (0.001)/((4/3)(3.1415)(0.152) ) = 0.0608 m

Run the Mie theory algorithm:
Mie(nr,
x)
--->

Qs

Mie(1.1806, 2.0097) ---> 0.1971

Calculate the scattering coefficient:
2

s = QsA = (0.1971)(0.0726) = 0.01431 m

s = s s = (0.0608)(0.01431) = 0.0009723 m-1
s [cm-1] = (s [m-1])(104 [m/cm]) = 9.723 cm-1

2.5.5 Scattering versus wavelength for 3 particle sizes

Consider the three optical scattering properties, s, g, and s(1 - g), as functions of wavelength for a
spherical particle with np = 1.5721 in a medium with nmed = 1.3316 at a concentration of 0.1% volume
fraction (fv = 0.001). Assume that np and nmed are constant versus wavelength for this example
calculation.
Let the sphere be one of three sizes: 0.100 m, 0.300 m, and 1.00 m.
The Mie theory calculation yields:

s(1 - g)

Note: The true np and nmed for polysytrene spheres and water are slightly wavelength dependent and
would deviate slightly from the above example.

2.6 Mie scattering from cellular structures

Soft tissue optics are dominated by the lipid content of the tissues. Such lipid constitutes the cellular
membranes, membrane folds, and membranous structures such as the mitochondria (about 0.5
m). While other objects such as protein aggregates and the nucleus are also sources of scattering,
the lipid/water interface of membranes presents a strong refractive index mismatch and so plays a
major role is scattering. The following graph summarizes the lipid contents of various tissues:
Lipid contents of tissues

[ref. 1] Click figure to enlarge

Mie theory provides a simple first approximation to the scattering of soft tissues [ref. 2]. The
approximation involves a few assumptions:
Assume the refractive index of the lipid membranes of cells is 1.46, based on the reported
1.43-1.49 range for hydrogenated fats [ref. 3].

Assume the refractive index of the cytosol of cells is 1.35, based on the reported value for
cellular cytoplasms [ref. 1].
Assume the lipid content of soft tissue is about 1-10% (f v = 0.02-0.10). Let's choose f v = 0.02
for this example to match the value for several typical soft tissues such as lungs, spleen,
prostate, ovary, intestine, liver, arteries, to name a few.
Assume all the lipid is packaged as small spheres of various sizes whose number density
maintains a constant volume fraction f v.
Ignore the interference of scattered fields from particles which can alter the apparent
scattering properties based on isolated particles.
In the following graphs, a 2% volume fraction of lipid is packaged as spheres of one size, for a series
of choices of sphere diameter (radius a), adjusting the number density to maintain a constant volume
fraction:
3
s = f v/((4/3) a )

s(1 - g)
This picture includes the experimental values of s (1 - g) for dog prostate,
illustrating that the prostate is mimiced by a 2% volume fraction of lipid spheres
in the 0.200-0.600 m diameter range. Other soft tissues vary only a little from
this general pattern for prostate tissue.

In summary
The wavelength dependence of light scattering in soft tissues such as the prostate suggests
that cellular structures equivalent to spheres with diameters in the range of 0.200-0.600 m contribute
most of the scattering. The magnitude of the scattering suggests that a volume fraction of 0.02 or 2%,
which is roughly the reported lipid content of prostate and many other soft tissues, yields a magnitude
for s(1 - g) which matches the magnitude of s(1 - g) for prostate. Deviations from our assumed
values for np based on lipid and nmed based on cytosol and from our assumed value for fv will affect
the magnitude of s(1 - g). But in general,soft tissues with higher (or lower) lipid content will show
increased (or decreased) scattering, while the wavelength dependence of s(1 - g) should not change
greatly.

2.7 Mie scattering from collagen fibers

Collagen fibers are strongly scattering. For example, the dermis of skin or the sclera of the eye are
tissues with high collagen fiber content. Mie theory can be used to approximate the scattering
properties of collagen on two levels:
on the macroscopic level of collagen fiber bundles.
on the ultrastructural level of periodic striations in collagen fibrils.
macroscopic level of collagen fiber bundles

Collagen fibers vary from 0.1 m-dia. fibrils to 8 m-dia. fiber bundles. A study reported the
distribution and concentration of collagen fiber bundle diameters in 9 post-mortem neonatal skin
specimens [ref.4, Saidi et al. 1995].
The fiber diameter (d) was 2.80 0.08 m (n = 9), i.e., (mean SD for n specimens), with an
intraspecimen variation SD/mean = 0.43 0.05.
The fiber concentration was s = 3 x106 0.5 x106 cm-3. The volume fraction was f v = d2(1
cm) s /4 = 0.21 0.10 (n = 9).
Mie theory can approximate collagen fiber scattering in dermis using the following assumptions:

Assume nmed of the dermal ground substance is 1.350, based on a reported value for the n of
corneal ground substance [ref.5].
Assume the np of the collagen fiber bundles is 1.389, based on a mean dermal water content
of 65.3%, W = 0.653, (see table 9.1 of ref.1): approximately, n = 1.50 - (1.50 - 1.33)W = 1.389.
Assume one can use the cylindrical Mie theory calculation of Bohren and Huffman [appendix
C in ref.6], which assumes that the incident light is oriented perpendicular to the long axis of
the fiber cylinders.
Assume that each fiber cylinder scatters light independently, ignoring any interference
effects from closely spaced fibers.
The above assumptions allow calculation of s(1 - g) versus wavelength for dermal collagen fiber
bundles. (see "Mie" in figure below).

ultrastructural level of periodic striations in collagen fibrils

The ultrastructure of collagen fibrils presents periodic striations as was shown in the figures in
our introduction to scattering. Fibrils are composed of entwined tropocollagen molecules, presenting
a banded pattern of periodic striations (70-nm spacing) due to the staggered alignment of the
tropocollagen molecules which each have an electron-dense head group that appears dark in the
electron micrograph. The periodic fluctuations in refractive index on this ultrastructureal level appear
to contribute a Mie scattering component that dominates the visible and ultraviolet wavelength
ranges. Such scattering from very small structures is called the Rayleigh limit of Mie scattering, or
simply "Rayleigh" scattering.
The following figure compares Mie theory for various size spheres withthe "Rayleigh" component
of skin scattering seen experimentally. The "Rayleigh" behavior of skin scattering is mimiced by 50nm-dia. spheres, np = 1.5, nmed = 1.35, at the volume fraction of collagen in dermis, f v = 0.21. This
assignment of the "Rayleigh" scattering to the collagen ultrastructure should be regarded as only a
working hypothesis, but there is no other material in large quantity in the dermis to offer a strong
source of scattering.

Click on figure to enlarge

Combining the Mie and Rayleigh contributions for skin

The combination of the "Mie" and "Rayleigh" contributions to scattering are shown in the following
graph, along with measured skin data:
"Mie" contribution is Mie scattering from
2.8-m-dia. cylindrical collagen fiber bundles,
np = 1.46,
nmed
=
1.35,
fv =
0.21.
"Rayleigh" contribution is Mie scattering from
50-nm-dia. spheres mimicing the ultrastructure
associated with the periodic striations of collagen fibrils,
np = 1.50, nmed = 1.35, f v = 0.21.
Click on figure to enlarge.