Color Formulation In Maxillofacial Elastomer By Genetic Algorithm

This work proposed to develop a portable, accessible, and objective solution for the reproduction of colors in medical-grade silicone elastomer by means of genetic algorithm and smartphones. The smartphones' application for color recipe prediction followed the two-constant Kubelka-Munk theory with corrections of internal and external reflection by Duntley, Saunderson and Richmond. The calculation of the composition of the mixture of pigments (n = 9) and flocking (n = 11) was performed by applying the genetic algorithm. Then, an intuitive application was developed for clinical use. The user sets the thickness and weight of the prosthesis and chooses a background option. The results screen displays the top five formulations generated by the genetic algorithm and their respective estimated color differences based on the CIELAB (Delta E) system under two illuminants (average sunlight and incandescent light). Simulations of color formulation were carried out and the estimated means of color difference were below the thresholds accepted clinically (Delta E < 3), but above the threshold of human visual perception for incandescent light (Delta E >= 1.1). So, different modifications on the algorithm were tested and the mean color difference was significantly improved (Delta E < 1.1). Though the mean value under average sunlight increased, it remained under the perceptive threshold. The methodology showed encouraging results and points out the potential use of smartphones to assist the clinician during the color formulation and to reduce costs.