关键词: 年龄推断, 颈椎骨龄, 牙龄, 多元回归分析, 维吾尔族

Abstract: PURPOSE: To explore the correlation between the actual age, dental age and cervical vertebral bone age in Uygur children and adolescents in Xinjiang region, and to construct the regression equation and joint prediction model of cervical vertebral bone age for males and females to improve the accuracy of age estimation. METHODS: This study was conducted in two phases. In the first stage, the data of full mouth surface tomography and skull localization lateral radiographs of 320 Uygur children and adolescents aged 8-15 years were collected. The dental age was inferred by Willems method, and the third and fourth cervical vertebrae were measured by Mito cervical vertebrae osteometry. The cervical spine bone age measurement data of 224 samples (112 males and 112 females) were randomly selected for multiple stepwise regression analysis. The cervical spine bone age prediction equations for males and females were established respectively, and verified with the remaining 96 samples (48 males and 48 females). In the second stage, the cervical spine age of all samples was calculated based on the regression equation constructed in the first stage. Subsequently, Spearman correlation coefficient was used to analyze the correlation between chronological age, dental age and cervical spine bone age, and further multiple regression analysis was performed to evaluate the predictive effect of dental age and cervical spine bone age on chronological age. Finally, all the sample data were collected and combined multiple regression analysis was carried out to construct a comprehensive age prediction model. RESULTS: The cervical spine bone age equation established through multiple stepwise regression analysis was: male cervical spine bone age=-1.189+16.607×H4/D4+9.064×AP3/H3+5.369×H4/AH4(R=0.875); female cervical spine bone age=0.233+7.084×AH4/AP4+26.048×h4/H4+4.650×AH3/PH3(R=0.868). Spearman correlation analysis results showed that the correlation coefficient between chronological age and dental age was r=0.956, the correlation coefficient between chronological age and cervical spine bone age was r=0.871, and the correlation coefficient between dental age and cervical spine bone age was r=0.843, all of which showed significant positive correlation(P<0.01). The unified regression model constructed by the joint analysis further improved the prediction accuracy, and the R² of the model was 0.916. The final formula of the combined model was: actual age=-1.064+0.753× tooth age +0.325× cervical bone age. CONCLUSIONS: Through joint analysis of dental age and cervical bone age, this study significantly improves the prediction accuracy of actual age, and successfully constructs a unified prediction model suitable for Uyghur children and adolescents aged 8-15 years in Xinjiang region, providing an efficient and reliable tool for age inference in the region.

Key words: Age inference, Cervical bone age, Dental age, Multiple regression analysis, Uyghur