Harmonizing color measurements in dentistry using translucent tooth-colored materials

Abstract

Abstract Background Since color measurements are relative, the discrepancy among different instruments is alarmingly high. This multicenter study evaluated the effectiveness of instrument calibration and inter-instrument harmonization of different spectrophotometers with the same optical geometry using tooth-colored, translucent dental materials. Methods The coordinating center (CC) spectrophotometer was calibrated using the NPL Ceram Series II set. Two sets of 10 specimens, labeled 1 to 10 and I to X (10 mm in diameter and 1 mm thick), were tested at CC and three research sites (RS1, RS2, and RS3) using the same d/8° optical geometry spectrophotometers. Calibration factors were calculated for each material and site to obtain the average calibration factors for sets 1–10, set I-X, and the combination of both. The differences among the non-corrected and corrected reflection values were calculated using CIEDE2000 (DeltaE00) and CIELAB (DeltaEab) color difference formulas and were submitted to ANOVA and Tukey test (α = 0.05). Results A significant decrease of color differences between non-corrected as compared to corrected measurements was recorded for all CC-RS and RS-RS comparisons. The reduction of DeltaE00 values between non-corrected and corrected for CC-RS1, CC-RS2, and CC-RS3 were 83.1%, 77.2%, and 73.6%, respectively. The corresponding DeltaE00 values for RS1-RS2, RS1-RS3, and RS2-RS3 comparisons, indirectly compared in the experiment, were 84.2%, 82.8%, and 68.5%, respectively. There was a significant reduction of DeltaE00 and DeltaEab color difference for all combined RS pairs and each of three RS pairs, corrected with one of two specimen sets calibration factors separately. Conclusions Calibration and harmonization of color measurements in dentistry using tooth-colored, translucent restorative materials significantly decreased measurement discrepancies between the coordinating center and research sites and among pairs of research sites.