Page 263 - Vibrational Spectroscopic Imaging for Biomedical Applications

Raman Micr oscopy for Biomedical Applications 239

8.2.4 Human Lung Tissue
Raman and FTIR imaging were applied to study the spectral features
26
and derive molecular information of normal lung tissue, congenital
27
cystic adenomatoid malformation (CCAM), and bronchopulmo-
28
nary sequestration (BPS). Other Raman work in the context of lung
have been reported for the detection and grading of neoplasia in cell
29
samples and for the characterization of normal bronchial tissue. 30
CCAMs are benign masses of nonfunctional lung tissue developing
during the embryonic gestation from an overgrowth of the terminal
bronchioles with subsequent suppressing alveolar growth. BPS
involves abnormal pulmonary tissue lacking trancheobronchial con-
nection and having a separate vascular supply that arises from the
thoracic and abdominal aorta. Cryosections were mounted on CaF
2
substrates. Raman images were collected at a step size of 66 to 100 μm
to assess the whole tissue section. The step size was reduced to 10 μm
to resolve details in selected areas. Datasets were segmented by k-
means cluster analysis and the averaged spectra of cluster were com-
pared. Photomicrographs and Raman microscopic images of normal
lung tissue (Fig. 8.7b) and congenital cystic adenomatous malforma-
tion (Fig. 8.7c) are compared. Normal lung tissue shows the typical
sponge-like morphology. Morphological changes are visible in solid-
type CCAM. Here, the tissue sections have fewer holes which are
consistent with nonfunctional, nonaerating, abnormal masses of lung
tissue. A cluster analysis of the combined Raman images distin-
guishes six tissue classes. Normal lung tissue is dominated by the
classes represented by red and green clusters. The averaged Raman
spectrum is shown in Fig. 8.7a as trace (A). CCAM is dominated by
the tissue class represented by the blue cluster. The averaged Raman
spectrum is shown in Fig. 8.7a as trace (C). Three additional tissue
classes in CCAM were assigned to smooth muscles of vessels (yellow,
olive) and microcrystals (black). The Raman spectrum of these micro-
crystals indicated a high content of the phospholipid phosphatidyl-
choline. Whereas it was not possible to identify the lipid type using
infrared spectroscopy, cluster analysis of FTIR images revealed a fur-
ther tissue type with spectral contributions of glycogen. This confirms
a general tendency that components having high content of polar
C—O, O—H and C—H groups such as glycogen show usually stron-
ger infrared than Raman bands. Furthermore, these results clearly
demonstrate that Raman and FTIR complement each other.
A typical Raman spectrum of BPS is included in Fig. 8.7a as trace
(B). As compared with Fig. 8.4 and 8.5, main differences between the
Raman spectra are assigned to spectral contributions of hemoglobin
that decreases in the order normal lung tissue > BPS > CCAM. As
described in more detail in Sec. 8.3.2 for red blood cells, Raman bands
of the heme group are enhanced by a resonance effect. On the one
hand, this enables to detect changes in the state or content of hemoglo-
bin more sensitive than using infrared spectroscopy. On the other hand,