Comparative projectile trauma: an examination of the differences in skeletal trauma inflicted by firearms and archery weapons
Koppenhaver-Astrom, Courtney Fiona
In recent years, the defining characteristics of cranial projectile trauma have been reported extensively in experimental studies as well as forensic case reports. The existing literature, however, focuses on trauma inflicted by firearms, primarily handguns and rifles. Though firearms are the most common form of projectile weapon used in a forensic context, there are several types of projectile weapons which have not been examined through experimental research. This gap in the literature not only limits the examination of forensic cases, but also inhibits the examination of trauma found within an archaeological context. This study sought to differentiate the skeletal trauma caused by different projectile weapons that are classified as either firearms (handgun, rifle, and shotgun) or archery weapons (recurve hand bow with field tip arrows, compound hand bow with fixed broadhead arrows, and compound crossbow with field tip bolts, fixed broadhead bolts, and mechanical broadhead bolts). Using polyurethane spheres as proxies for human cranial vaults, samples were shot by one of the specified weapons (n=5) and 35 features resulting from projectile impact (both qualitative and quantitative) of the entrance and exit defects were recorded. Using principal component analysis, it was found that the features of trauma which accounted for the highest proportion of variance observed in the subset which included both entry and exit defects were the maximum fracture length on the external table of the entrance site, the minimum fracture length on the external table of the entrance site, the entrance defect diameter, the minimum fragment length of the fragments that originated from the entrance defect, the width of the reconstructed exit defect, the maximum fracture length on the external table of the entrance defect, and the width of the reconstructed entrance defect. These accounted for 96.74% of the variance within this dataset. When only examining the entrance defects, the most distinguishing variables were the maximum fracture length on the external table of the entrance defect, the width of the entrance defect, the minimum fracture length on the external table of the entrance defect, and the width of the reconstructed entrance defect, accounting for 95.89% of the variance within this dataset. Machine learning (linear discriminant analysis) was applied to test the predictive strength of these variables. In testing the accuracy of these predictions, it was found that the program could correctly predict the weapon used for 74.19% of the samples when examining both the entrance and exit defects and 60.87% of the samples when only examining the features of the entrance defect. The findings of this research exhibit the indiscernible qualitative features between trauma inflicted by different projectile weapons, calling to attention the need to change the current methods of weapon identification. This study has established new quantitative methods for projectile trauma analysis which are simple to perform, require minimal equipment, and are easily applied to forensic and archaeological remains.