Journal of Stomatology

Abstract

1/2024 vol. 77
Original paper

Forensic sex classification using cut-off values from cranio-mandibular radiologic parameters

  1. Faculdade de Medicina Dentária Universidade de Lisboa, Portugal
  2. Medicina Forense, Faculdade de Medicina Universidade de Lisboa, Portugal
  3. Forensic Analysis Group, Centre of Statistics and its Applications (CEAUL), University of Lisbon, Portugal
  4. FORENSEMED, Research Group from UICOB Unit, Faduldade de Medicina Dentária da Universidade de Lisboa, Portugal
  5. School of Technology and Management, Polytechnic Institute of Leiria, Portugal
J Stoma 2024; 77, 1: 14-20
Online publish date: 2024/02/29
View full text
Confronting perimenopausal women’s knowledge of coronary heart disease with their health behaviours. Controversial role of hormone replacement therapy in the protection of coronary heart disease

Introduction

In the context of forensic identification, sex determination is the first important step, and the mandible is of particular interest because it shows clear signs of dimorphism. OBJECTIVES: The purpose of this investigation was sex classification through the application of several cranio- mandibular metric variables in a Portuguese adult population.

Material and methods

206 orthopantomographic images (116 females and 90 males) were selected from Portuguese adults aged between 25 and 79 years. From each image, 14 linear variables were measured using ImageJ® software. IBM SPSS® software version 27.0 was applied to analyze data presented as extremes, mean, standard deviation, and median. Statistical methodologies included intra-class analysis, normality test, t-test, Mann-Whitney test, logistic regression, and ROC curve to determine cut-off point and respective sensitivity and specificity of sex classification.

Results

Comparing the two sexes, all variables showed significant differences, except for the angular variables. Variables with the best area under ROC curve were height of the right (0.861) and left (0.850) mandibular ramus. Based on the logistic regression, the left coronoid height, right mandible body height, left mandible body height, and right mandible ramus height showed an accuracy level of 83.0%, sensitivity of 86.2%, and specificity of 78.9% in sex classification.

Conclusions

The obtained results show identifiable mandibular dimorphism in the measurements of mandi­bular structures in orthopantomography images. These metric variables allow for the creation of discrimination models between males and females in a Portuguese adult population.

Share