Color prediction of layered dental resin composites with varying thickness

The aim of this study was to assess the accuracy of a Principal Components Analysis (PCA)-based method for reflectance reconstruction and color estimation of layered dental resin-based composites with different thicknesses.

Bi-layered samples of different clinically relevant thicknesses were created using shades of VITAPAN Excell (VE), VITAPAN Dentine (VD) and VITA Physiodens (VP), combined with their corresponding enamel shades. Spectral reflectance of all samples was measured over a black background using a non-contact spectroradiometer with CIE 45^∘∕0^∘ geometry. Two different PCA-based models, built from two different configurations of known samples, were proposed to reconstruct the spectral data and color of unknown layered samples. Root Mean Square Error (RMSE), Goodness of Fit (GFC), as well as ΔE₀₀ with corresponding 50:50% acceptability and perceptibly thresholds (AT and PT) were used as performance assessment.

The 5-samples training set approach provided an average RMSE < 0.015 and GFC > 0.999 when measured and predicted spectral reflectances were compared, while for the 9-samples training set, RMSE < 0.0098 and GFC > 0.9999 were obtained. The overall mean color differences obtained with the 5-samples training set approach were ΔE₀₀ = 0.99 (AT% = 96.25% and PT% = 32.50%), while using the 9-samples training set resulted in lower overall mean color differences: ΔE₀₀ = 0.50 (AT% = 99.22% and PT% = 83.87%).

Within the framework of this study, the two proposed PCA-based configurations allow the prediction of the spectral reflectance of layered dental resin-based composites of different shades and thicknesses, with a high degree of accuracy.