Essay On Regression Analysis

Age, one of the essential factors in establishing the identity of persons, can be assessed by inspection of non-invasive radiographic imaging of teeth. Alternative approaches such as non-invasive radiographic techniques were used to study the amount of secondary dentine deposited by measuring the pulp chamber size reduction. This reduction in pulp chamber was correlated with chronological age and regression equations were derived to estimate the age.
Aims and Objectives:
Our aim was to estimate age in people of coastal Andhra region by deriving independent regression equation by measuring the pulp chamber crown-root trunk height ratio using digitalised orthopantamographs of mandibular 1st molars.

Materials and Methods:
Orthopantomographs

There is a high negative correlation (r = - 0.745) between PCTHR and Age and it is statistically significant p<0.01.
Regression analysis on the study group gives rise to the following equation for age estimation: “Estimated Age = -125.8 × PCTHR + 53.77” (Derived from study subset).
Standard deviation of PCTHR related to age (each age group) was given in Table 4, fig 4 and standard deviation of PCTHR related to gender was given in fig 5.
This equation was applied on the test group to test its accuracy in estimating the age. No significant difference was found between the actual age and calculated age (P =0.1889). MAE was estimated by applying the equation on the test sample and was found to be 1.73 years. There was no significant difference between the actual age and calculated age in males (P =0.4575) and females (P = 0.2521).

Result for total 100 test subjects:
There is no significant difference between actual age and estimated age (by using regression equation) in test group. Z-test is used to test significance and p>0.05
Result for total 50 male test