A study of 3-D CT scan geometric morphometrics for classification of human crania in Malaysian population

Abstract

Estimation of ancestry and sex are the main steps for identification of unknown skeletonized human remains. The present study aimed to use geometric morphometricanalysis for the classification of Malaysian crania to different ancestries and sexes. A total of 390 3D-CT scans of postmortem crania (130 Chinese, 115 Indian, 145 Malaywere retrieved from Forensic Department in Hospital Kuala Lumpur and UKM Medical Centre. Two approaches were used. In the first approach, about 19 landmarks representing the face and base were located in 3Dcoordinates on the retrieved data using the Stratovan Checkpoint. The second approach showed about eight landmarks around the mastoid were located and exported in morphologika format. The analysis was done by using MorphoJ software, and the visualization was done by using Landmark Editor Software. The first three principal component analysis accounted for 20.2%, 13.5% and 9.3% of the total shape variance, respectively in the first approach. Variations among groups were estimated by canonical variate analysis and discriminant function analysis. There was high classification accuracy between different groups up to 96.1 % between Chinese and Indian. All variations between sexes of each ancestral group were significant, which was approved by permutation test. Procrustes analysis of variance was used to estimate differences between the crania according to size and shape. Sex showed high variation in size, while ancestry showed high variation in shape. The visualization resulted in different variations between crania. In brief, geometric morphometric proved to show significant visualization of different sex and ancestry. The results can be used for database collection for the Malaysian population.