Find in Library

Objective To establish a mathematical model of stature estimation for Sichuan Han females on the basis of the relationship between lower limbs and individual height, thus to provide evidence for forensic identification. Methods Samples were collected from 171 Sichuan Han females. Large flat panel multi-function digital photography system was used to take the full-body X-ray films of the lower limbs. Indexes of long bones and stature of the subjects were measured, respectively. A linear regression analysis was carried out on the correlation between them, and a mathematical model of the stature calculation was established. Then the mathematical model was used to calculate the stature of another 29 Sichuan Han females to test its accuracy. Results The maximum length of femur （x1） had the highest correlation with stature. A total of 13 linear regression equations were established （P<0.05）, with the correlation coefficient （R） 0.821-0.897 and the standard error of the estimation （SEE） 2.994-3.812 cm. The backtesting showed that the equation y=41.604+1.205 x1+1.318 x6+2.444 x12+1.852 x13-2.388 x14 had the smallest mean absolute deviation （2.485 years old） and the highest accuracy of ±2SEE （92.9%）, and that the equation y=48.783+2.568 x1 had the highest accuracy of ±1SEE （60.7%）. Conclusion The stature estimation is high by using the long bones of the lower limbs has high accuracy.