| |
The aim of this study was to analyse visible spectral data from a range of early caries lesions. Sixty naturally occurring, non-cavitated, early caries lesions on the approximal surfaces of extracted teeth were divided into three equal-sized groups; group 1 (dull white-spot lesions), group 2 (shiny white-spot lesions) and group 3 (brown lesions). The teeth were exposed to a 405 nm laser diode light source (Nichia Ltd.) and spectral intensity data were collected with a 1.5 mm diameter probe from the centre of the lesion and an adjacent area of sound tooth using a fibre-optically-coupled spectrometer (Avantes Inc.). This allowed accurate and repeatable placement of the probe tip over the areas of interest (sound and carious tooth). Spectral intensity data between 450 nm and 750 nm were analysed. Initially the intensity data were normalised and plots of the intensity difference between sound and lesion at each wavelength interval were made for each tooth. This initial data processing allowed further analysis of the differences between the lesion types using the un-normalised data. Four 20 nm wavelength bands were selected (480-500, 550-570, 580-600, 620-640 nm). For each tooth, ratios of the intensity in each wavelength band were calculated for both the sound and lesion spectra. The analysis demonstrated average differences between brown- and white-spot lesions of around 50% and between dull and shiny lesions of around 12%. The technique may potentially provide a quantitative method of distinguishing between lesion types. Supported by EPSRC and the British Council. |