Size and shape variations of the human scapula bone using a geometric morphometric method for sex estimation

Sex estimation is one of the crucial biological profiles required for the identification of unknown skeletal remains. Using the geometric morphometric approach, this study aimed to assess the size and shape differences of three-dimensional (3D) scapula models between the sexes. Twelve three-dimensional landmarks that represent the shape of the scapula were digitized on 470 of the 3D scapula models. Following generalized Procrustes analysis, centroid size and Procrustes coordinates were used for the size and shape analysis, respectively. In the size analysis, there was a significant difference between male and female scapula models with a cross-validation classification accuracy of 90.6%. For shape analysis, the principal component analysis revealed 29 principal components (PC) with PC1 to PC6 displaying more than 5% of variation for each PC. Using one-way analysis of variance (ANOVA) and multivariate analysis of variance (MANOVA), PC1 to PC5 showed significant shape differences between males and females. Discriminant function analysis of the shape data revealed 80.3% of cross-validation classification accuracy. In conclusion, the 3D scapula models showed significant size and shape differences between sexes with good classification accuracy. This may contribute to the development of population-specific databases in the Malaysian population and benefits the forensic teams.