Duangto Phuwadon, Janhom Apirum, Prasitwattanaseree Sukon, Mahakkanukrauh Pasuk, Iamaroon Anak
Graduate Program in Forensic Osteology, Faculty of Medicine, Chiang Mai University, Chiang Mai 50200, Thailand; Forensic Osteology Research Center (FORC), Faculty of Medicine, Chiang Mai University, Chiang Mai 50200, Thailand.
Forensic Osteology Research Center (FORC), Faculty of Medicine, Chiang Mai University, Chiang Mai 50200, Thailand; Department of Oral Biology and Diagnostic Sciences, Faculty of Dentistry, Chiang Mai University, Chiang Mai 50200, Thailand.
Forensic Sci Int. 2016 Sep;266:583.e1-583.e5. doi: 10.1016/j.forsciint.2016.05.005. Epub 2016 May 13.
The aims of this study were to develop new prediction models for dental age estimation and to test the accuracy of the resulting models in comparison with the Demirjian et al. and the Willems et al. methods in Thai children and adolescents. Digital panoramic radiographs of 1,134 Thai individuals (487 males and 647 females) aged from 6 to 15 years were selected and evaluated for dental age estimation. Quadratic regression was used to generate new models. The results showed that the new prediction models indicated a strong correlation coefficient between the dental maturity score and the chronological age in both sexes (r=0.951 for males, r=0.945 for females). The new age prediction models were: y=0.006297x(2) - 0.804930x+32.591843 for males and y=0.010677x(2) - 1.538823x+61.955056 for females, where y is the dental age, x is the dental maturity score according to Demirjian et al.
Moreover, these new models were tested showing the greatest accuracy for estimating the age in Thai samples using the mean difference values between the dental and the chronological ages (-0.04 years for males, 0.02 years for females) when compared with the Demirjian et al. and the Willems et al.
In addition, the new models revealed a high percentage of accuracy in the absolute difference values between the dental and the chronological ages within 1 year (76.26% and 74.49% for males and females, respectively). Furthermore, our results in mean difference values indicated that the Demirjian et al. method (0.11 and 0.10 years for males and females, respectively) was more accurate than the Willems et al. method (-0.37 and -0.39 years for males and females, respectively) in Thai samples. In conclusion, the new age prediction models in this study provide accurate age estimation in both sexes, suggesting that these models be applied for forensic age estimation, especially in Thai children and adolescents.
本研究的目的是开发用于牙龄估计的新预测模型，并与德米尔坚等人和威廉姆斯等人的方法相比，测试所得模型在泰国儿童和青少年中的准确性。选取了1134名年龄在6至15岁的泰国个体（487名男性和647名女性）的数字化全景X光片，并对其进行牙龄估计评估。使用二次回归生成新模型。结果表明，新预测模型显示出牙成熟度得分与男女实际年龄之间具有很强的相关系数（男性r = 0.951，女性r = 0.945）。新的年龄预测模型为：男性y = 0.006297x² - 0.804930x + 32.591843，女性y = 0.010677x² - 1.538823x + 61.955056，其中y为牙龄，x为根据德米尔坚等人的方法得出的牙成熟度得分。
此外，对这些新模型进行了测试，结果显示，与德米尔坚等人和威廉姆斯等人的方法相比，使用牙龄与实际年龄之间的平均差值（男性为-0.04岁，女性为0.02岁）时，这些新模型在估计泰国样本年龄方面具有最高的准确性。
此外，新模型在牙龄与实际年龄相差1年内的绝对差值方面显示出较高的准确率（男性为76.26%，女性为74.49%）。此外，我们的平均差值结果表明，在泰国样本中，德米尔坚等人的方法（男性和女性分别为0.11岁和0.10岁）比威廉姆斯等人的方法（男性和女性分别为-0.37岁和-0.39岁）更准确。总之，本研究中的新年龄预测模型在男女中均能提供准确的年龄估计，表明这些模型可应用于法医年龄估计，尤其是在泰国儿童和青少年中。
