前牙美学修复智能选色设计系统的初步研究

目的 旨在建立前牙修复体人工神经网络智能选色设计系统，提高前牙瓷修复体的美学质量。方法 通过数码相机采集图像，图像预处理；采用一种基于三刺激值的相对选色方法及21点平滑牙色过渡算法的九分法的牙齿颜色匹配；牙齿立体感的恢复；建立牙色库，对样本牙、缺失牙的牙色特征选择、提取，采用人工神经网络智能技术进行前牙修复体颜色的计算机自动选择设计和个性牙色设计。结果 基于人工神经网络的前牙瓷修复体颜色的计算机自动匹配选择具有较高的准确性，通过计算机辅助设计的病例，经同期随机对比，患者、医生、观众满意度均明显高于未经过计算机辅助设计者，患者对修复体满意度的评价比医生的评价高。结论 前牙修复体人工神经网络智能选色设计系统有较大的临床应用价值，该研究促进了人工神经网络技术在口腔修复领域的应用。

Intelligentized design of anterior prosthesis for color selection

Objective To set up a neural network model and system to be applied in color selection of anterior porcelain prosthesis, and to improve its cosmetic effect. Methods The image was automatically captured by digtal camera, and pre processed. The values based on the three stimulation value were adopted for the colour selection and tooth colour matches from nine zone colors with 21 point smoothly transitional calculation. The three dimension view of the teeth was recovered. Tooth colors storage was set up and the characteristics for sample teeth and lost teeth were selected and retrieved to match the colour through artificial neural network automatically for personalized design of teeth colour. Results The method of anterior prosthesis for computerized color selection based on artificial neural network had a higher accuracy. By comparision, the satisfaction of patients, dentists and viewers was much higher than that without computer aided design, and patients' eveluation for the prosthesis was higher than that of dentists. Conclusion The design system for colour selection of the front tooth prosthesis has a great value in clinical application, and it has developed the theory of artificial neural network used in the field of stomatology.