Sex estimation in Indians by digital analysis of the gonial angle on lateral cephalographs

Abstract

Objective: Sex estimation of skeletons is important in forensic reconstructive identification. The mandible is a durable component that is suitable to discriminate the sexes while lateral cephalometry is a standardised radiographic technique accepted as a tool in personal identification. Limited data is available for the mandibular/gonial angle as a parameter for sex assessment using lateral cephalometric radiographs. The aim of this study is to determine the gonial angle’s accuracy in sexing Indians using a new digital method and statistical approach.
Method: The sample comprised of 304 digital lateral cephalometric radiographs (155 females and 149 males, age between 18-30 years) of Indian subjects. The mandibular/gonial angle was measured on these radiographs using Adobe Photoshop software using tools available therein. The obtained angles for the sexes were subjected to logistic regression analysis (LRA), which forms a composite of weighted independent variables using a multivariate strategy.
Results: The average angle was 122.7° for females and 121.1°for males. LRA produced an accuracy rate of 56.3% in sex assessment, with females being more accurately identified (61.9%) than males (50.3%).
Conclusion: The study demonstrated significant univariate sexual dimorphism among males and females in this population. However, the sex prediction value of this approach was low and thus may not be useful in sex estimation involved in human identification of Indians.