Three-dimensional prediction of nasal growth using landmark-based morphometry: Clinical validation of a predictive...

Area of Science:

• Orthodontics and craniofacial surgery
• Medical imaging and computational anatomy
• Adolescent growth and development

Background:

• Predicting adolescent nasal growth is crucial for facial treatment but current methods lack precision.
• Existing three-dimensional (3D) facial imaging models often lack sufficient nasal resolution for accurate growth prediction.
• This study validates a novel 3D nasal growth prediction algorithm using dense landmark-based morphometry.

Purpose of the Study:

• To clinically validate the geometric accuracy of a 3D nasal growth prediction algorithm.
• To assess the algorithm's performance in predicting post-pubertal nasal morphology.
• To evaluate the algorithm's clinical applicability in adolescent nasal growth assessment.

Main Methods:

• Developed a predictive model using a cross-sectional dataset of 2050 multi-ethnic subjects (ages 6-19) with 54 nasal landmarks.
• Trained a supervised multivariable linear regression model on morphometric parameters, age, and sex.
• Externally validated the algorithm on a separate cohort of 12 longitudinally followed adolescents, comparing predicted vs. observed morphology.

Main Results:

• The algorithm demonstrated sub-millimetric accuracy, with mean surface deviation of 0.58 ± 0.20 mm.
• 95.6% of nasal surface points were within a 1.5-mm clinical threshold.
• The algorithm showed high concordance for Goode ratio and nasal tip projection, outperforming a no-growth baseline.

Conclusions:

• The validated 3D algorithm offers high geometric accuracy for nasal growth prediction in adolescents.
• The findings support the algorithm's clinical utility for treatment planning and patient counseling.
• Further refinement with larger prospective datasets will enhance precision across diverse populations.