Objectives: This investigation aims to demonstrate the potential to differentiate sex by examining the morphology of the mandible, using three-dimensional statistical shape modelling.
Methods: Mandible data obtained from CT scans of each 44 deceased Korean male and female adults was post-processed to concentrate solely on morphology. Subsequent analysis of processed data revealed morphological differences between the average male and female mandible models, which were then graphically visualized using vector data, and statistical significance was assessed.
Results: The overall size of the mandibles did not significantly vary between males and females. However, male mandibles were characterized by a more medially developed mandibular condyle and coronoid process, contributing to a narrower total width in comparison to females. Also, males displayed lateral growth at the gonial angle and upper development at the extramolar sulcus. Conversely, the female mandible had a greater anterior-posterior length. Statistically significant differences (p<0.05) were found in the gonial angle region and extramolar sulcus. A classification learner in MATLAB using an Efficient Linear Support Vector Machine achieved the best accuracy at 76.1%. With this model, ANOVA and Kruskal Wallis tests identified Principal Components 49 and 12 as key in determining sex, representing changes in the coronoid process and the mandibular condyle and gonial angle, respectively.
Conclusion: The findings from this study provide significant insights into the sexual differences in human mandible morphology and underscore the potential of statistical shape modeling in forensic anthropology. Furthermore, this research paves the way for future development of a classification model based on mandible morphology.
Keywords: mandible; sexual dimorphism; statistical shape modeling
Ethical statement: The data (Digital Korean) used in this research complies with all applicable ethical standards. The data is sourced from open, public database, thus ensuring the anonymity and privacy of individuals. Due to the nature of the data, this study did not require formal ethical approval as it does not involve any direct interactions with human subjects or manipulation of confidential data.
Funding statement: This work is supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) in 2023 (NRF-2018R1A6A7023490). Acknowledgement: The presenter gratefully acknowledges the human data support provided by Korea Institute of Science & Technology Information (KISTI) which produced these data with the Catholic University of Korea.