Biomedical Engineering Reference
In-Depth Information
quartz
window
flow in
liquid core waveguide
ordinary tubing
flow out
common-path
epi-Raman
Fig. 16.3.
Schematic of a flow-through Raman-based biofluid analysis system. A
waveguiding tube delivers the fluid near a quartz window, through which laser light
is coupled into the waveguide to perform the spectroscopy [5]. The liquid then flows
through a groove to another piece of tubing that carries it away. The waveguiding
nature of the delivery tube creates a long optical sampling region, thereby enhancing
the Raman signal
collect Raman emission through the tube walls with abutting fiber arrays. The
technique exploits the index mismatch at the quartz/air interface to achieve
waveguiding conditions, confining the light to travel down the tube, and excite
the whole sample eciently [4]. Another waveguiding-tube approach, used by
Qi et al. and depicted in Fig. 16.3, is to use a common-path geometry, both
delivering and collecting light from one entrance to the tube. In this case, the
waveguiding not only steers excitation light down the tube but also funnels
some of the Raman emission back to the entrance aperture, thereby enhancing
the collection eciency. By placing an optical window near the entrance to
the tube and allowing the liquid to enter via a right-angle bend, this approach
is compatible with a sealed-off liquid flow-through process [5].
In Vivo
The most intriguing “container” for biofluids is, of course, the body itself.
While in vivo measurements of urine hold no appeal because it is regularly
and painlessly excreted, blood measurements are obviously attractive. There
have been three main approaches for transcutaneous Raman spectroscopy of
blood, all sketched in Fig. 16.4.
The first is to acquire a “pure blood” spectrum by volume localization.
Confocal microscopy can routinely interrogate regions more than 100
m
below the tissue surface, which is deep enough to reach blood capillaries. Be-
cause light from out-of-focus regions is strongly rejected, the method makes it
possible to gather a more or less “pure blood” spectrum without an interfer-
ence signal from surrounding tissue structures (Fig. 16.4a).
Another tactic is to compute a “difference spectrum” contrasting two mea-
surements of a tissue region taken before and after some change in the blood
content, as sketched in Fig. 16.4b. This provides a spectrum that is nominally
μ
Search WWH ::
Custom Search