Morphological analysis of frontal sinus for human identification amongst Malaysian: a skull radiograph and computed tomography study

Biological profiling is necessary for unknown skeletonised remains. In cases of fragmented skeletons, the resiliency of the frontal sinus (FS) suggests its potential in establishing biological identity. Osteometry method is limited to external skeletal evaluation, whereas imaging allows for internal structural analysis. Given the population-specific nature of forensic identification, studies on FS morphology for biological profiling is least being studied and yet to be explored among Malaysian population. Thus, determining suitable identification methods via imaging, to analyse the morphology of FS in Malaysians is necessary. This study aimed to evaluate FS pattern distribution and determine morphological variations by sex, race and age profiles among Malaysians using descriptive analysis, two-dimensional (2D) geometric morphometric, three-dimensional (3D) volumetric, linear measurements and anterior wall thickness analyses respectively. Samples of 504 skull radiographs and 398 head computed tomography (CT) scans of adult Malaysians comprising males and females of Malay, Chinese and Indian races aged of ≥20 years old, were retrieved. The FS patterns in posteroanterior skull radiographs were descriptively classified according to the presence or absence of FS. Eight 2D landmarks were applied to the lateral skull radiographs and 2D geometric morphometric analysis was performed using MorphoJ. Three-dimensional models of the FS and its anterior wall were reconstructed from CT scans. Volumetric and linear parameters were determined for the 3D FS. Five types of grids with 12 regions of interest (ROIs) were developed and applied onto the 3D FS anterior wall area. The thickness of the anterior wall within the 12 ROIs were recorded. Results showed that bilateral presence of FS is common (95.4%) and bilateral absence is rare (2.7%). Procrustes ANOVA revealed significant differences in FS shape by sex (p<0.0001) and race (p=0.0205), with 80.6% and 57.4% classification rates, respectively. However, no significant shape differences were observed within age groups. FS volumes differed significantly by sex (p=0.000), race (p=0.035), and age group (p=0.039), with classification rates of 68.0%, 33.1%, and 37.4%, respectively. Linear measurements also showed significant output by sex (p=0.000) and race group (p=0.001) with accuracies of 80.2% and 41.5%, while no significant differences were noted for age. Anterior wall thickness exhibited significant variations by sex (p=0.001), race (p=0.036), and age group (p=0.011), with classification rates of 62.5%, 44.8%, and 40.2%, respectively. Overall, the rarity of FS absence suggests its applicability in forensic identification. Both 2D geometric morphometric on lateral FS, and linear measurement on 3D reconstructed FS model are reliable for sex identification among Malaysians. This population-specific study could assist in enhancing the FS database within the Malaysian context and provides a valuable reference for forensic identification based on FS morphology.