3D- Facial Analysis: Principal Component Analysis of Facial Soft Tissue Morphological Features in Three Adult Population

Objective:

This preliminary study seeks to identify variations in soft tissue facial morphology among ethnically diverse, facially balanced adults through Principal Component Analysis (PCA) of three-dimensional (3D) facial scans.

Materials and Methods:

A total of 210 3D facial scans (35 individuals per subgroup: Chinese, Hungarian, and Hispanic; stratified by sex) were selected from the University of Alabama’s database. Scans were obtained using 3D laser scanning and stereophotogrammetry (3dMD). Fifty-seven anatomical landmarks were manually identified on each scan. Landmark coordinates were processed using Generalized Procrustes Analysis and PCA (performed in R) to determine the principal components most responsible for overall shape variation.

Results:

PCA revealed four principal components (PCs) explaining 77.72% of the total variance in soft tissue morphology:

PC1 (49.13%) – Associated with upper facial height.

PC2 (17.70%) – Represented the relationship between nasal protrusion and eye position.

PC3 (6.31%) – Reflected interocular distance and vertical eye alignment.

PC4 (4.14%) – Corresponded to upper lip projection.

The analysis indicated that the greatest morphological variability occurred in the upper facial region. These findings were interpreted in both clinical and aesthetic contexts and aligned with previous research on facial growth and attractiveness.

Conclusion:

Notable differences in upper facial soft tissue morphology were observed among ethnically distinct but balanced adult groups. The study underscores the effectiveness of PCA in uncovering clinically relevant facial shape patterns and emphasizes the value of personalized, culturally informed approaches in orthodontic and orthognathic treatment planning. Future research incorporating longitudinal data and AI-based methods is encouraged to enhance personalized diagnostics and establish inclusive aesthetic benchmarks.