Raman Spectroscopy of Bone and Cartilage - Emerging Raman Applications and Techniques in Biomedical and Pharmaceutical Fields - page 352

higher crystallinity. While further work in this area is warranted, the results
are significant because both crystallinity and mineral/matrix ratio may be
measurable non-invasively, as discussed below.

14.8 Subsurface Raman Spectroscopy and Imaging

Prior to true subsurface bone spectroscopy, Penel and coworkers obtained bone
Raman spectra using a titanium chamber with a fused silica window placed in
the calvaria of New Zealand rabbits [2]. With this apparatus they were able
to study both bone tissue and implanted hydroxyapatite and

-tricalcium
phosphate over a 8-month period. In addition to bone spectra, hemoglobin
spectra were obtained close to blood vessels.
The first subsurface bone tissue Raman spectroscopic measurements were
performed using picosecond time-resolved Raman spectroscopy on excised
equine cortical bone [56, 57]. In these experiments it was shown that a
polystyrene backing could be detected through 0.3 mm of bone. The same
picosecond technology was used to perform the first transcutaneous Raman
spectroscopic measurements of bone tissue [58]. In this study, the cortical bone
mineral/matrix ratios of excised limbs of wild type and transgenic (oim/oim)
mice were compared and the differences demonstrated.
Transcutaneous Raman spectroscopic measurements using spatially offset
optical fibers were reported less than a year later [59, 60]. The test systems
were chicken tibiae and the humeri of human cadavers. The use of cadaveric
and ex vivo specimens allowed validation of the measurements by comparison
to exposed bone tissue. In these measurements a depth of 3-4 mm below the
skin was reached. In vivo measurements began with a report of the Raman
spectrum of a phalange of a human volunteer [61]. The periosteal surface was
probably 1-2 mm below the skin and the mineral phosphate

β

ν 1 was accurately
reproduced, although incomplete separation of mineral and matrix spectra
introduced errors in other bands.
Using a ring-disk probe (also called inverse SORS) in backscattered ge-
ometry we have been able to reach bone about 6 mm below the skin [60].
Mineral/matrix ratios are accurately ( < 4% relative error) reproduced if a
protocol for optimizing the ring/disk spacing is followed [62].
Very recently, ex vivo diffuse Raman tomography of canine bone tissue
has been reported [63]. Importantly, signal was collected even through tissue
thicknesses of 40 mm (Fig. 14.6). This encouraging result suggests that non-
invasive in vivo bone Raman spectroscopy at clinically significant depths may
be feasible. In diffuse optical tomography an image is constructed from the
data obtained with excitation and collection fibers placed at different points
on a specimen. The reconstruction is an ill-posed inverse problem and best
results are obtained if the shape of the object is known independently. In an
extension of that work, it has been shown that high contrast can be obtained