Page 254 - Vibrational Spectroscopic Imaging for Biomedical Applications
P. 254
230 Cha pte r Ei g h t
2852 2886
1442
1439 1298 1661 (A) 717 875 1066 1089 1130 1270 1301 1660 1738 (D)
Raman Intensity 427 544 608 700 716 1004 1064 1128 1268 (B) 1400 1592
1004
1243 1340 1447 1665 2932 457 599 727 976 1124 (E)
508 621 643 760 828 853 940 (C)
400 700 1000 1300 1600 3000 3300 700 1000 1300 1600 3000 3300
–1
–1
Raman Shift (cm ) Raman Shift (cm )
−1
−1
FIGURE 8.2 Raman spectra from 400 to 1800 cm and 2700 to 3550 cm from
fi ber tracts (A, orange cluster in Fig. 8.1), normal tissue (B, blue/cyan cluster in
Fig. 8.1) of mouse brain, a protein mixture (C), the lipid phosphatidylcholin (D) and
metastasis of malignant melanoma in mouse brain (E, black cluster in Fig. 8.1).
The Raman spectrum of fiber tract shows the typical signature of
white matter with maximum spectral contributions of lipids and cho-
lesterol. Spectral contributions of cholesterol are evident at 427, 544,
−1
608, and 700 cm . The intensity ratio of lipid-to-protein bands
decreases in Raman spectra of gray matter which is represented by the
blue and cyan cluster in normal mouse brain tissue. Raman spectra of
brain metastases can easily be distinguished by intense spectral con-
tributions of the pigment melanin at 457, 599, 727, 976, 1124, 1400, and
−1
1592 cm . The pigment partially absorbs the excitation wavelength
785 nm which gives a preresonance effect in the spectral range 400 to
−1
1800 cm . It is important to note that the expression of the pigment
melanin is a molecular property of the primary tumor cells. That means
the secondary tumor contains the molecular information of the pri-
mary tumor. As Raman spectroscopy probes the molecular fingerprint
of tissues and cells, this opens the exciting perspective to develop a
Raman-based approach for identification of the primary tumor of
metastases. Classification models have already been reported for the
related vibrational spectroscopic technique infrared spectroscopy.
They assigned the primary tumors of the four most frequent human
brain metastases based on the infrared spectroscopic fingerprint. 10,11
Unknown primary tumors constitute a severe problem in patients with
brain metastases as an organ specific therapy cannot be performed.
The fiber optic probe which was used in this proof-of-concept
study combines several attractive features. Integrated filters suppress
the elastic Rayleigh scattering and background signals that are gener-
ated by the intense excitation laser inside the fiber. A lens focuses the
laser light onto the sample and enables optimized collection geometry