Determination of human sex by odontometric characteristics using equations of logistic regression

Abstract

BACKGROUND: The development of methods for identifying a person based on bone remains, which are adapted for diagnosing the population of Azerbaijan, is a priority in view of the incompleteness of solving this problem. In this situation, a new way to quickly determine the sex by teeth appears to be in demand because it should increase the efficiency of the work of forensic doctors. AIM: To develop a modern technique for diagnosing a persons sex using odontometric signs. MATERIALS AND METHODS: Teeth sizes were studied at the craniological collection of Azerbaijanis. On the seven teeth of the upper and seven teeth of the lower jaw, three parameters were measured (while symmetrical teeth were considered a single array). Each of the 14 anatomical teeth was represented by 80 male and 80 female teeth. A digital caliper (Mutitouyo; accuracy 10 microns) was used for odontometric procedures. Odontometry results were analyzed by statistical methods using receiver operating characteristics curve analysis. Mathematical modeling was conducted using the MATLAB statistical software package. RESULTS: With the help of the receiver operating characteristic curve (ROC) analysis, the data obtained in the course of odontometry were studied. Six odontometric signs were related to sex. The most sex-related parameters were the buccolingual diameter of the upper canine, heights of the crowns of the second lower molar, lower incisors, lower canine, and upper central incisor. Using mathematical modeling, six logistic regression equations have been implemented, which help determining the sex of an individual by the size of his teeth possible. The predictive potential of the equations was also assessed using the ROC analysis. CONCLUSION: Based on odontometric indicators and ROC analysis, diagnostic equations for determining the sex of a person by the teeth size were obtained. The equations have not yet been verified on independent samples; therefore, they cannot yet be unconditionally recommended for practice. Nevertheless, the authors count on the high legitimacy of their models because they used a fairly powerful mathematical apparatus.