DentShadeAI: a Framework for Automatic Dental Shade Matching through Mobile Phone Camera

The accuracy of the current shade matching method is affected by clinician experience and the condition of light sources; and the standard shade matching equipment is expensive and may not be available in remote areas. In this work, we propose an artificial intelligence framework, DentShadeAI, which predicts the closest dental shade, VITA classical A1-D4, to the target tooth via an image from a mobile phone camera. The framework consists of an image capturing procedure and the machine learning process. We designed the capturing procedure so that the photo-taking can be done through a mobile phone camera without using controlled light sources. In addition, we evaluated three machine learning models: Random Forests, SVM, and XGBoost. The result showed that the Random Forest model outperformed the other models; the best model achieved 97.1% accuracy and 97.2% F1-score.