Biomedical Engineering Reference
In-Depth Information
There are, however, limits to this method. First, assuming typical variation in experimental data, the
SHG intensity versus pinhole size calibration curves may not be able to distinguish between
F
i
/
B
i
values
much above ~5 (note the decreasing separation of the curves in Figure 17.6, as
F
/
B
⨠5). Fortunately,
F
/
B
ratios (extrapolated to
F
i
/
B
i
as appropriate, for comparison) in many biological samples of interest,
including cancer tissues, are often <5 [26−29,48,57]. Second, because this model assumes there is no
subsequent scattering of initially backscattered SHG
B
i
, this model is limited to tissue imaging depths
<1 MFP, such that significant scattering of
B
i
will not occur.
17.2 Polarization of SHG in tumors
17.2.1 Polarization Properties of collagen SHG
Collagen SHG emission is influenced by the polarization of the incoming excitation photons, together
with the overall orientation of the scattering collagen fibril relative to the laser axis, and the pitch angle
of the collagen helix (see the Appendix). Consequently, polarization of SHG signals can be exploited to
interrogate these tissue properties, as well as related properties such as collagen organization and align-
ment. In this section, we review the efforts to determine collagen fibril orientation and helix pitch angle,
and their application to understand the structure of tumor stroma. Before discussing the application of
these SHG polarization-related techniques to tumor tissues (to our knowledge, there have only been a
handful of such reports), it is informative to discuss some of the preceding literature in nontumor tissue
that has helped to lay the foundation for interpreting polarization effects of SHG as they apply to col-
lagen's structure, organization, and molecular properties.
The work from several groups has helped us to define and understand the relationships between
incoming laser polarization and collagen SHG emission intensity [27,36,37,39,40,58-60], and these
relationships are in turn principally sensitive to the orientation of SHG-emitting dipoles [60]. As
such, SHG polarization anisotropy (PA) involves rotating the polarization plane of the incoming lin-
early polarized laser light relative to the plane of the SHG emitter, and measuring the resultant SHG
intensity (
I
SHG
) over the range of relative polarization states. In general,
I
SHG
minima and maxima
occur when laser polarization is perpendicular and parallel to the collagen fiber direction, respec-
tively [61,62]. Of particular importance for imaging clinical samples, to help prevent sample damage
that may occur by taking scans at many different polarization states, methods have been advanced
to allow for obtaining meaningful PA data from as little as four scans [39], or from continuous scan-
ning techniques [40]. In addition, this dependence of
I
SHG
on laser polarization is impacted by tissue
depth [40,51], proportionate to tissue scattering effects that will effectively depolarize the laser emis-
sion before it reaches the focal volume [51]. This phenomena should be heeded by investigators as
they compare anisotropy measurements at different tissue depths and across tissues with different
scattering properties, and indeed has been addressed in part by an elegant “optical clearing” method
of tissue preparation that serves to reduce scattering effects in biologic tissues, thus normalizing the
polarization dependence at varying tissue depths [51]. Thus, scattering effects relative to tissue thick-
ness and the location of the focal plane are also important considerations when deciding whether to
measure the forward or backward SHG signal, both of which have been employed for polarization
anisotropy.
For the purposes of our discussions in this chapter, PA relationships with
I
SHG
enable mathemati-
cal derivation of several key pieces of information related to collagen's structure and organization.
First, we can obtain the angular orientation of an SHG-emitting collagen fibril in the XY plane
(see the Appendix, Figure A.1), information that in turn can be used to generate “orientation field
maps” reflecting the overall orientation of fibril or fiber ensembles throughout an field of view (FOV)
[38,39]. Second, we can determine a related “anisotropy parameter” β, which represents a measure of
collagen ordering in the sample, and is calculated by: β = (
I
SHGpar
-
I
SHGperp
)/(
I
SHGpar
+ 2
I
SHGperp
), where
