遗传算法预测不同错类型牙冠宽度

Prediction about Mesiodistal Crown Diameters of Teeth from Different Classes of Malocclusion Based on Genetic Algorithms Method

目的:使用遗传算法建立针对不同错类型的各牙牙冠宽度与其它牙的相关方程,并将同类型同性别的各牙用一个公式进行表达。方法:选取安氏Ⅰ、Ⅱ、Ⅲ类患者初诊模型各100副,每种类型男女各50副,45副作为模拟样本,5副作为检验样本。用游标卡尺测量各牙冠宽度,对不同错类型分别采用遗传算法(Genetic Algorithms,简称GAS)建立各牙与其他牙冠宽度的关系方程,并与线性回归分析法进行精度比较。结果:建立了不同错类型各牙与其他牙冠宽度的相关方程。GAS法与线性回归分析法的预测值分别与实测值比较,绝大多数牙位无统计学意义,但GAS法的检验样本平均误差绝对值均小于线性回归分析法。结论:GAS法的预测精度优于线性回归分析法,建立的方程有利于对不同错类型牙列萌出不全患者进行全牙列间隙分析。

Objective：To establish an equation between mesiodistal crown diameters of different types of malocclusion teeth and other teeth,and to use just one equation to describe the situation of the teeth with same position and same gender.Methods：Pre-treatment casts of malocclusion patients＇ diagnosed as classⅠ,Ⅱ,Ⅲ were screened out.100 for each type,and 50 males 50females respectively.Among these,45 were designed as analog samples and the other 5casts were set as test samples.An equation for these three classes of malocclusion was established between the mesiodistal teeth diameter measured by a vernier caliper and that of the other teeth,and then compared with linear regression analysis.Results：The equation between mesiodistal crown diameters of different types of malocclusion teeth and that of other teeth was established.Comparing the predictive value with measured value,there existed no statistical significance neither GAS nor linear regression analysis but the absolute mean error of test samples by GAS was less than by linear regression analysis.Conclusion：The prediction accuracy by GAS is better than by linear regression analysis method.The established equation is of value to do full dentition gap analysis for different types of malocclusion.