Infrared and Raman Spectroscopy and Imaging of Breast Cancer - Vibrational Spectroscopy for Tissue Analysis - page 138

• Note the distorted (sigmoidal) bands in the relectance spectra,
particularly visible as the dip in reflectance to a higher wavenum-
ber of the amide I band at ~1639 cm −1 . In theory, this effect should
shift the band maxima to a lower wavenumber (by a few tens of
wavenumbers), but this does not seem to be the case here, as the
positions are very similar to the ATR positions. The amide II peak
(~15 3 5 c m −1 ) is shifted to a lower wavenumber, as expected. In the
KKT spectra, the amide II peak is in a position similar to the ATR
spectra.
• Overall, it can be stated that the KKT transform does not make a
consistent improvement in the spectra, so the best option is to use
the reflectance spectra.
• There are substantial differences between the ATR and relectance
spectra at wavenumbers lower than about 1350 cm −1 .

Analysis of Individual Images

There are a number of methods that can be employed to analyse individual
and complex images. In this section, these methods are described using the
samples described in Table 5.4.

Single-Wavelength Images

To reduce the effects of baselines, the second derivative of the spectra can be
been taken. One way to visualise the images is to do false-colour plots based
on the intensity at particular wavenumbers corresponding to peak maxima
(which become minima in second-derivative spectra).
Using one of the spectrums as an example, nine peaks can be chosen from
the mean spectrum (Figure 5.20). The second derivative intensity at these

Mean for 94-73261a-g3

× 10 -4

2

0

-2

-4

-6

1000 1200 1400 1600 1800

Wavenumber/cm -1

Figure 5.20
Peaks selected for false-colour images.