Biomechanically Based Correlate for Localized Lung Contusion From Nonlethal Blunt Impact Projectiles.

INTRODUCTION Injury mechanics of blunt impact projectiles differ from those experienced in whole body motor vehicle collisions because the effects are localized around the point of impact, and thus, injury thresholds based upon gross chest kinematics (e.g., force, velocity) may not be applicable across impact types. Therefore, knowledge of biomechanically based tissue injury correlates for blunt impact projectiles are needed to better guide design and development of protective systems as well as assess injury risks from blunt impact projectile weapons. MATERIALS AND METHODS In this study, subject-specific swine finite element models were used to quantify the tissue-level stresses and strains resulting from high speed projectile impact. These tissue-level injury doses were correlated to pathology injury outcomes to produce injury risk curves for lung contusion. Details of the pathology data and finite element results are provided in Appendix 1. Survival analysis regression methods were applied to develop lung injury regression curves and a number of statistical methods were used to evaluate several biomechanical metrics as correlates to lung contusion. Uncertainty and sensitivity analyses were used to further confirm the selection of the correlate. RESULTS Statistical analysis revealed that normalized strain-energy density was the best correlate for prediction of lung tissue damage. Going further, normalized strain-energy density also proved to be suitable for prediction of the percentage of contused lung volume, a more meaningful medical diagnosis. As expected, peak strain-energy density is most sensitive to muscle-skin properties, as quantified through a comprehensive uncertainty and sensitivity analysis over three sets of projectile weights and speeds. CONCLUSIONS Normalized strain-energy density was found to be the best correlate for prediction of lung tissue damage and correlate well to extent of contused lung volume.