Characterisation of Deep Layers of Tissue and Powders: Spatially Offset Raman and Transmission Raman Spectroscopy - Emerging Raman Applications and Techniques in Biomedical and Pharmaceutical Fields - page 59

et al. [51] who (using Kerr gating) successfully detected the spectral signa-
ture of osteogenesis imperfecta (brittle bone disease) in mouse limbs through
1 mm of overlying soft tissue. The presence of the condition was apparent
from differences in relative band intensities between the organic (collagen)
and mineral (phosphate) components. This represented a major milestone,
although instrumental complexity and the excessively high laser intensities
(two to three orders of magnitude above the legal limits for skin illumination)
precluded an extension to human subjects.
These issues were subsequently addressed with the advent of SORS, which
only requires continuous wave laser beams with substantially lower inten-
sities. The first use of SORS in the area of non-invasive spectroscopy of
bones was reported by Schulmerich et al. [30] who obtained Raman spectra of
bone from depths of several millimetres in animal and human cadavers using
multivariate curve resolution (band-targeted entropy minimisation (BTEM)).
The subsequent use of a ring illumination geometry (equivalent to inverse
SORS) by Schulmerich et al. [35] permitted further increases in the quality
of Raman spectra and penetration depth. The team succeeded in determin-

ing the ratio of the intensities of the phosphate 958 cm − 1 and carbonate

1070 cm − 1 bands of chicken tibia (a potential indicator of the presence of
osteoporosis [52]), through 4 mm of overlying soft tissue to within an accu-
racy of 8%. Typical Raman spectra obtained in these experiments are shown in
Fig. 3.10.
This group has developed methodology for in vivo subsurface spectroscopy
of anaesthised mice and other small animals [53]. In this study of 32 animals,
they demonstrated that there is no statistically significant difference between
measurements made in vivo and on the exposed cortical bone of the femur
after killing of the animals. Because of the small size of the mouse, alignment of
the probe over the bone is critical for obtaining accurate results. It is expected
that it may be easier to do in vivo measurements in larger animals because of
the greater ease of positioning multifibre probes.
Parallel research by Matousek et al. [32] demonstrated the feasibility of ob-
taining the Raman spectra of bones from humans in vivo. Although the Raman
spectra were of limited quality with some overlying tissue signals present, the
experiment demonstrated that the key spectral features of human bone could
be measured transcutaneously in vivo under safe illumination conditions. The
measurements were performed with a laser power of only 2 mW, approximately
an order of magnitude below the safe level for skin illumination in the near
infrared. In these measurements a ring fibre probe with 0 and 3 mm spatial
offsets was used; scaled subtraction of the spectra obtained at the two offsets
produced an estimate for the spectrum of the underlying bone. The measure-
ments were performed on the thumb distal phalanx bone of volunteers through
2 mm of overlying soft tissue. For further details of the research in this area
the reader is referred to [Chap. 14].