Data-driven standards for infant skull thickness distributions in computational modeling and analysis

Area of Science:

• Biomechanical Engineering
• Pediatric Traumatology
• Computational Anatomy

Background:

• Child abuse is a critical global issue, with infants under one year facing the highest fatality risk.
• Computational modeling aids in predicting injury and validating histories to prevent abuse, but lacks anatomical variability data.
• Accurate injury prediction requires understanding natural anatomical variations in populations.

Purpose of the Study:

• To quantify skull thickness distributions in infants to establish data-driven anatomical standards.
• To enhance the accuracy of computational models for predicting head trauma injuries in infants.
• To provide age- and sex-based guidelines for injury prediction models.

Main Methods:

• Quantified skull thickness in 266 infants, analyzing age and head circumference as predictors.
• Categorized infants under 12 months into four age groups using thickness distribution breaks and variance optimization.
• Assessed sex differences and lateral symmetry in skull thickness across developmental stages.

Main Results:

• Head circumference was a better predictor of skull thickness than age for infants under 2.5 months.
• Identified four distinct age-based skull thickness categories for infants under 12 months.
• Found no significant average sex differences in skull thickness, but noted 53 specific locations with differences; lateral symmetry is a reasonable assumption for infants under 12 months.

Conclusions:

• This study establishes the first data-driven categorization of infant skull thickness distributions.
• Generated essential guidelines for age- and sex-based computational models in predicting infant head trauma.
• Findings contribute to improved injury prediction accuracy and prevention of child abuse through enhanced biomechanical analysis.