A quantitative assessment of Saw Kerf floor shape patterns using outline analysis

In dismemberment cases, forensic anthropologists evaluate bony cut surfaces and estimate saw class characteristics, which can aid in investigative and legal proceedings. Previous publications indicate that saw class characteristics, such as tooth shape, saw set, and power, can be deduced from the kerf profile shape and size. However, these studies are based on subjective visual categorizations, at times with limited statistical assessments. This study used elliptical Fourier analysis to quantitatively assess relationships between kerf shapes and saw class characteristics. Incomplete kerf profiles (n = 133) made with 19 saws in anatomically gifted, macerated human limbs (n = 19) were assessed. Kerf profiles were captured with a stereomicroscope and closed outlines were created and subjected to elliptical Fourier and principal component analyses. PerMANOVAs and Kruskal-Wallis analyses were performed on the resultant principal components to assess the effects of saw set, power, and tooth shape on kerf shape. Cross-validated stepwise discriminant function analyses (DFA) were performed to evaluate classification accuracy. There was no significant difference in entrance and exit morphology (p = 0.31). Significant results were obtained for all saw class characteristics. DFA classified tooth shape with 88.0 % accuracy. Flat and U-shaped kerfs were associated with rip saws while W-shaped kerfs were indicative of crosscut saws. DFA classified saw power with 89.5 % accuracy. On average, mechanical saws produced kerfs with larger widths compared to hand saws. Relationships between kerf floor morphology and saw set, however, were more complex. These quantitative analyses of kerf shape generally support anecdotal relationships established in the literature and its utility in forensic assessment.