Methods: Extracted bovine teeth were cleaned and covered with acid resistant nail varnish so as to leave a window of exposed enamel. The teeth were then immersed in orange juice (pH 3.4) for up to 36 hours after which time they were examined by a Proscan™ 2000 non-contact surface profilometer. Five different methods of measuring the depth of eroded lesions in the teeth were evaluated and compared; 2-point band (step height), 2-point square (step height), 3-point square (step height), extrapolation of an intact tooth surface (difference) and a selecting a region of interest within the extrapolated volume (difference).
Results: Whilst there was generally good agreement between the different techniques used, some general trends were observed. The 2-point band (step height) method yielded the deepest lesion depth values whereas extrapolation of the tooth surface (difference) gave the shallowest. ANOVA showed that the variance between different teeth using a single depth measuring method was the same as that when using the five different methods upon the same tooth. Accurate superimposition of scans taken before and after acid immersion was problematic; however, scans taken at a higher resolution, lower scan rate and with repeated sampling will alleviate these difficulties.
Conclusion: The natural curvature of the teeth manifests itself in the different depth measurement techniques and steps must be taken to minimise this distorting effect when evaluating mineral loss. Whilst the superimposition of a longitudinal set of scans remains the paradigm for evaluating mineral loss in teeth, a more rapid turnover of data is possible using the 2-point band step height method.