Benson, Wyatt H., authorPante, Michael C., advisorDu, Andrew, committee memberJackson, Jessica, committee member2023-06-012023-06-012023https://hdl.handle.net/10217/236554Forensic methods must adhere to the Daubert standard to be deemed as admissible evidence in court. Current critiques regarding how well this standard is upheld have also challenged whether current forensic practices truly meet the Daubert standard. For example, kerf mark analyses can reveal trace evidence in sharp force trauma cases but a lack of quantitative studies and standardized analytical methods leave the field open to potential scrutiny. While previous research frequently classifies marks as either the product of serrated or non-serrated blades, further identifications are rarely made confidently. The goal of this project is to determine whether variations in 3D micromorphological variables can be used to quantitatively discriminate between kerf marks made by different knife types and blade classes. Here, kerf marks were produced using five different knives on bovid diaphyses, 3D scanned using profilometric microscopy, measured for both volumetric and profile variables, then analyzed using quadratic discriminant analysis. Results show individual knives were classified correctly in only 52% of attempts. However, blade class – serrated vs. non-serrated vs. partially serrated – was successfully identified in 97% of attempts. Significantly, our results differentiate between kerfs produced by serrated blades, non-serrated blades, and partially serrated blades, not only allowing for more specific blade identifications but also producing a quantifiable and replicable method meeting the Daubert criteria.born digitalmasters thesesengCopyright and other restrictions may apply. User is responsible for compliance with all applicable laws. For information about copyright law, please see https://libguides.colostate.edu/copyright.bone surface modificationkerf markstaphonomyforensicsanthropologysharp force traumaCut it out: a novel, quantifiable approach to kerf mark analysis using 3D confocal microscopy and machine learningText