Dental age estimation of Malaysian juveniles using population-specific standards / Safar Sumit Bunyarit

Dental age (DA) of an individual is estimated by comparing the stage of dental growth against published reference dental surveys. Numerous methods of tooth maturity assessment in age prediction have been described. One method introduced by Demirjian et al. (1973), hereafter known as the original method, and another one improvised by Chaillet & Demirjian (2004), hereafter known as the modified method, are two of those most widely employed as standards. In this retrospective cross-sectional study, both methods were applied to Malaysian juveniles; when the methods were found to be unsuitable, a population-specific method for DA estimation was developed based on formation of the left mandibular permanent teeth. A total of 4614 dental panoramic tomograms of Malaysian juveniles of known chronological age (CA) comprising 2301 males and 2313 females aged 5-18 years old were examined. The DA was compared with the CA using paired t-test. The mean age of an individual tooth for each developmental stage was calculated; the DA was converted using Demirjian’s maturity scores. Additionally, the pattern and level of tooth development within dental arches were determined and compared by age, gender and ethnicity. The dental maturity of Malaysian juveniles was generally overestimated when Demirjian’s original method was used. Malay boys and girls were overestimated by 0.36±0.93 and 0.25±0.83 years, Chinese boys and girls by 0.47±0.86 and 0.34±1.01 years, and Indian boys and girls by 0.56±0.97 and 0.43±0.90 years, respectively. In contrast, dental maturity of these juveniles was underestimated when the modified method was applied. Malay boys and girls were underestimated by -2.09±0.90 and - 2.79±0.99 years, Chinese boys and girls by -1.92±0.94 and -2.76±1.05 years, and Indian boys and girls by -1.68±1.00 and -2.56±1 years, respectively. v Considering the over- and under-estimation of DA in the respective methods, Demirjian’s scores were adapted by using artificial neural networks (ANN) to produce more accurate DA estimation. With the ANN treatment, differences between the CA and DA when compared against Demirjian’s original method were statistically insignificant: for Malay boys and girls the differences were 0.05±0.83 and 0.04±0.79 years, for Chinese boys and girls, 0.002±0.78 and 0.028±1.11 years, and for Indian boys and girls, 0.039±0.86 and 0.032±0.93 years, respectively. Malaysian girls were generally faster than boys in tooth development especially in the first (Q1) and second (Q2) quadrants. The third molar mineralisation was similar between boys and girls, except for difference by one stage. Within dental arches, the third (Q3) and fourth (Q4) quadrants were more developed than Q1 and Q2. The difference between Q1 and Q2 and between Q3 and Q4 were statistically insignificant, whereas the difference between both Q1 and Q2 with Q3 and Q4 were significant for both genders and all ethnicities. Overall, Indians were more advanced compared to the Chinese, who in turn were more advanced than Malays. In conclusion, both the original and modified methods for DA estimation were inapplicable for Malaysian juveniles. Thus, a population-specific method was developed that improved the accuracy in DA estimation. The improved scores would be useful in clinical and forensic applications in the future