451
15 FtIR Imaging of tissues: techniques and methods of Analysis
with proteins (1404 cm
−1
) and their phosphorylation [105]. more detailed modelling
of the hPv effect on IR spectral profile of cell lines has been reported by ostrowska
et al. [106].
In a more recent study the effect of mie-scattering on FtIR images cervical
tissue was investigated [17]. Figure 15.8a shows uncorrected raw spectra selected
from regions containing superficial exfoliating cells (yellow), intermediate epithelium (pink), basal layer (yellow) and stroma (blue). Figure 15.8b shows the same
spectra after Rmie-EmSC. the Rmie-EmSC spectra are devoid of sinusoidal baseline modulation and the positions and intensities of the amide I mode are in the
expected range for these cell types. Figure 15.8c shows the h&E section of the
normal cervical epithelium; while Fig. 15.8d shows the corresponding amide I map
showing protein concentration after Rmie-scatter correction. Figure 15.8e shows a
map derived from uhCA performed on the uncorrected spectra, while Fig. 15.8f
shows the same map after Rmie-EmSC has been applied. the uhCA map produced from the RmieS corrected FtIR image is similar to the uncorrected result but
on closer inspection superior in that the epithelium shows improved layer demarcation into the exfoliating cell layer (gold), superficial layer of squamous epithelium
(dark green), the intermediate layer (pink and some glycogen rich areas in blue)
and the parabasal layer (orange). the stromal tissue is described by 3 clusters in
the uncorrected uhCA map but only 2 clusters in the corrected uhCA map. the
stroma is relatively homogeneous in terms of chemical content so the stroma being
described by fewer clusters correlates more satisfactorily with the known properties
of stromal tissue. these observations demonstrate that pre-processing FtIR image
spectra with the RmieS correction prior to performing uhCA can improve the clustering correspondence with tissue features as well as achieve clustering parsimony.
Although the studies presented in this section are promising, the diagnosis of
cervical cancer by the use of FtIR microspectroscopy is still a challenging task,
since many factors can affect biochemical, and consequently spectral status of this
tissue. Nevertheless, the potential application of FtIR spectroscopy to cervical cancer detection including cells and tissues is still an active research area [107–114].
most of these studies have focused on correlating the spectral information with
changes in biochemistry associated with the progression from healthy to malignant conditions and demonstrate the applicability of FtIR technique in tandem with
proper statistical tools for future cancer diagnostic systems.
15.5.2 Brain Tissue
Since FtIR microspectroscopic analysis requires tissue sections, most of the human brain samples originate from biopsy (only tumours) or autopsy. this limits
the applicability of FtIR imaging in diagnosis of diseases of the central nervous
system. despite this, the method has shown high sensitivity and specificity in grading brain tumours and its origin, even during surgery. here, spectral changes are
found in the entire mid-IR region of the spectrum revealing changes in intensity of
15 FtIR Imaging of tissues: techniques and methods of Analysis
with proteins (1404 cm
−1
) and their phosphorylation [105]. more detailed modelling
of the hPv effect on IR spectral profile of cell lines has been reported by ostrowska
et al. [106].
In a more recent study the effect of mie-scattering on FtIR images cervical
tissue was investigated [17]. Figure 15.8a shows uncorrected raw spectra selected
from regions containing superficial exfoliating cells (yellow), intermediate epithelium (pink), basal layer (yellow) and stroma (blue). Figure 15.8b shows the same
spectra after Rmie-EmSC. the Rmie-EmSC spectra are devoid of sinusoidal baseline modulation and the positions and intensities of the amide I mode are in the
expected range for these cell types. Figure 15.8c shows the h&E section of the
normal cervical epithelium; while Fig. 15.8d shows the corresponding amide I map
showing protein concentration after Rmie-scatter correction. Figure 15.8e shows a
map derived from uhCA performed on the uncorrected spectra, while Fig. 15.8f
shows the same map after Rmie-EmSC has been applied. the uhCA map produced from the RmieS corrected FtIR image is similar to the uncorrected result but
on closer inspection superior in that the epithelium shows improved layer demarcation into the exfoliating cell layer (gold), superficial layer of squamous epithelium
(dark green), the intermediate layer (pink and some glycogen rich areas in blue)
and the parabasal layer (orange). the stromal tissue is described by 3 clusters in
the uncorrected uhCA map but only 2 clusters in the corrected uhCA map. the
stroma is relatively homogeneous in terms of chemical content so the stroma being
described by fewer clusters correlates more satisfactorily with the known properties
of stromal tissue. these observations demonstrate that pre-processing FtIR image
spectra with the RmieS correction prior to performing uhCA can improve the clustering correspondence with tissue features as well as achieve clustering parsimony.
Although the studies presented in this section are promising, the diagnosis of
cervical cancer by the use of FtIR microspectroscopy is still a challenging task,
since many factors can affect biochemical, and consequently spectral status of this
tissue. Nevertheless, the potential application of FtIR spectroscopy to cervical cancer detection including cells and tissues is still an active research area [107–114].
most of these studies have focused on correlating the spectral information with
changes in biochemistry associated with the progression from healthy to malignant conditions and demonstrate the applicability of FtIR technique in tandem with
proper statistical tools for future cancer diagnostic systems.
15.5.2 Brain Tissue
Since FtIR microspectroscopic analysis requires tissue sections, most of the human brain samples originate from biopsy (only tumours) or autopsy. this limits
the applicability of FtIR imaging in diagnosis of diseases of the central nervous
system. despite this, the method has shown high sensitivity and specificity in grading brain tumours and its origin, even during surgery. here, spectral changes are
found in the entire mid-IR region of the spectrum revealing changes in intensity of