Multimodal Biomedical Imaging Systems - Biomedical Optical Imaging Technologies: Design and Applications - page 332

[ 42 ], photothermal radiometry [ 43 ], Raman spectroscopy [ 44 ], and multiphoton
microscopy [ 45 ]. All of these methods characterize one or more property changes
of the tooth structure. However, none of them can detect dental caries in early stage
with high sensitivity and specificity.
Reflectance imaging is the most widely used method in caries detection. Dentists
commonly examine the teeth visually or through an intraoral camera by capturing
the backscattered light from the tooth surface. One problem with conventional
reflectance imaging is the specular reflection on the tooth surface. This specular
reflection can cause some false positives and obscure surface features.
The fluorescence of teeth was discovered by Stubell in 1911 with UV excitation.
In the early 1980s, Alfano and Yao reported the first systematic investigation of
tooth fluorescence with different light excitations [ 46 , 47 ]. It was discovered that
by using light of wavelengths 350, 410, and 530 nm, the emission wavelength
peaks obtained for teeth were, respectively, 427, 480, and 580 nm. The peaks of
the emission spectra of carious regions were shifted to the red portion of the
spectrum. Also, Alfano and Yao reported that the relative intensity of light in the red
portion was greater for a carious region than for a sound region. Two fluorescence
techniques have been developed to detect dental caries. The first technology uses
blue light excitation [ 48 ]. When teeth are illuminated with high-intensity blue light,
namely, 400 nm, they emit light through the visible spectrum. Sound, healthy enamel
shows a higher fluorescence than demineralized enamel; demineralized areas appear
relatively darker under light that excites the fluorescence. The amount of mineral
loss in surface lesions correlates strongly with the loss of fluorescence signal.
Another technology developed by Hibst and Gall in 1998 uses red light excitation
[ 49 ]. When teeth are illuminated with red light, the fluorescence radiation from
healthy tooth regions is strongly reduced so that the fluorescence radiation from
carious regions is superposed with autofluorescence of the healthy tooth tissue.
Clean, healthy tooth structure exhibits little or no fluorescence. Red fluorescence
is proportional to the severity of tooth decay.
Dental OCT imaging was first reported in 1995 [ 42 ]. Preliminary work focused
on correlating optical scattering signatures with tissue structures in the periodontium
region of the oral cavity. Since then, several groups have investigated OCT imaging
for dental applications [ 50 - 52 ]. Amaechi et al. observed a linear correlation between
the mineral loss in enamel and dentine measured by transverse microradiography
(TMR) and the reflectivity loss measured by OCT. In 2005, Fried et al. demonstrated
that the increase in the integrated reflectivity in the PS-OCT perpendicular polar-
ization axis correlated well with increasing mineral loss/severity caused by natural
demineralization, by comparing with digital microradiographs (DM) and polarized
light microscopy (PLM).
While fluorescence imaging has good sensitivity in the detection of dental caries,
it doesn't provide information on lesion depth and suffers from low specificity
attributed to the presence of plaque, calculus, and stains on the tooth surface. In
addition, ambient light and hydration of tooth tissue also degrade the quality of the
fluorescence image. OCT imaging can provide high-resolution tooth images with