Biomedical Engineering Reference
In-Depth Information
[31,47,49,52,53]. In another recent study, Ajeti et al. [54] found that in mixed collagen V/collagen I gels,
the
F
d
/
B
d
ratio was effectively equivalent in 0% and 5% collagen V gels, but was significantly lower in 20%
collagen V gels. These results are significant because collagen V interacts with collagen I to influence
fibril diameter distribution [55] (0% and 5% collagen V gels have similar fibril diameter distributions
[55], hence no difference in
F
/
B
above), and collagen V is found upregulated in human breast carcinoma
[56]. Therefore, the ability to distinguish changes in proportions of collagen I/V (or other) isoptypes by
collagen
F
d
/
B
d
SHG may provide a useful prognostic tool for understanding and treating breast cancer.
Other groups have focused on developing novel instrumentation for obtaining
F
/
B
data from human
or other biologic samples. The most direct and typical means of obtaining
F
/
B
measurements involves
placing an objective lens on both sides of an excised and sectioned tissue specimen. However, this two-
lens approach is not feasible for very thick biopsy specimens, or for
in vivo
clinical applications such as
endoscopy.
To overcome these limitations, Han et al. [26] recently described a system that can capture collagen
F
/
B
SHG from intact specimens
in vivo
, using a single-objective lens. This system is illustrated in Figure
17.2a. Briefly, this system relies on the properties that at shallow imaging depths, the initially backward-
directed SHG (
B
i
) will exit the image from the focal volume with minimal subsequent scatter, whereas
the initially forward-propagating SHG (
F
i
) will subsequently scatter such that a fraction (as much as
~20%, see [57]) of this signal ultimately exits in the backward direction (and therefore can be captured
with the same objective lens and detector as the initially backward-directed SHG, i.e.,
B
i
) (Figure 17.2b).
This results in a Gaussian distribution of the
B
i
signal in an image plane, and a subsequently backscat-
tered
F
i
signal whose photon intensity distribution in an image plane is constant and does not vary
significantly with position over an approximately 50 μm radius from the focal volume [26] (Figure 17.3):
2
r
(17.1)
I
SHG
( )
r
=
B
exp
−
2
+
FC
ω
As defined in Ref. [26],
I
SHG
is the total SHG signal intensity distribution on the object plane,
r
is
the radius from the focal point on the image plane, ω is the e
-2
Gaussian spot size on the image plane
of the direct backward-propagating SHG (
B
i
), and
F
and
B
are the absolute intensities of the
F-
and
B-
propagating SHG signals. The
C
parameter relates the initially forward-propagating SHG (
F
i
) signal
intensity to the average intensity of the uniform distribution of (now backscattered) SHG light that
reaches the object plane, and is a function of the scattering and absorption properties of the tissue [26].
This equation can also be written as
2
r
F
B
C
I
SHG
( )
r
=
B
exp
−
2
+
(17.2)
ω
where
F
/
B
is the collagen fiber SHG
F
/
B
ratio [26] (i.e.,
F
i
/
B
i
, as we define it here). When a series of col-
lagen fiber SHG images are generated through a series of different sized confocal pinholes, each image
pixel represents an integration of the total SHG signal over the pinhole area that can be expressed as
follows:
2
2
π
R
r
+
F
B
∫
∫
(17.3)
I
∝
d
θ
exp
−
2
C
rdr
pixel
ω
0
0
where
R
is the pinhole size with respect to the direct backward-propagating SHG (
B
i
) Gaussian spot size,
that is,
R
=
r
pinhole
/ω [26]. By normalizing pixel intensities at the various pinhole sizes to the maximum
