Measurement uncertainty associated with chromatic confocal profilometry for 3D surface texture characterization of natural human enamel

- This paper reports a method for quantifying sources of measurement uncertainty following good metrology practice guidelines.
- The main measurement error was flatness deviations of 0.49 μm over 3 × 5 mm areas of an optical flat due mechanical movement of the scanning stage.
- Taking into account potential sources of error, an optimized measurement protocol was developed for 3D surface texture measurement of natural enamel undergoing dietary erosion.
- natural enamel samples showed significant reductions in 3D roughness after 45 min erosion (P < 0.05). This has clinical relevance for methods attempting to measure early erosion in vivo.

ObjectiveTo investigate the measurement performance of a chromatic confocal profilometer for quantification of surface texture of natural human enamel in vitro.MethodsContributions to the measurement uncertainty from all potential sources of measurement error using a chromatic confocal profilometer and surface metrology software were quantified using a series of surface metrology calibration artifacts and pre-worn enamel samples. The 3D surface texture analysis protocol was optimized across 0.04 mm2 of natural and unpolished enamel undergoing dietary acid erosion (pH 3.2, titratable acidity 41.3 mmol OH/L).ResultsFlatness deviations due to the x, y stage mechanical movement were the major contribution to the measurement uncertainty; with maximum Sz flatness errors of 0.49 μm. Whereas measurement noise; non-linearity's in x, y, z and enamel sample dimensional instability contributed minimal errors. The measurement errors were propagated into an uncertainty budget following a Type B uncertainty evaluation in order to calculate the Standard Combined Uncertainty (uc), which was ±0.28 μm. Statistically significant increases in the median (IQR) roughness (Sa) of the polished samples occurred after 15 (+0.17 (0.13) μm), 30 (+0.12 (0.09) μm) and 45 (+0.18 (0.15) μm) min of erosion (P < 0.001 vs. baseline). In contrast, natural unpolished enamel samples revealed a statistically significant decrease in Sa roughness of −0.14 (0.34) μm only after 45 min erosion (P < 0.05 s vs. baseline).SignificanceThe main contribution to measurement uncertainty using chromatic confocal profilometry was from flatness deviations however by optimizing measurement protocols the profilometer successfully characterized surface texture changes in enamel from erosive wear in vitro.