The influence of incompleteness and fragmentation of skeletal remains on novel methods for pair-matching is still unaddressed. For this study 154 innominate bones from 77 male individuals were selected. The bones presented different rates of preservation, which was systematically assessed, including three classes. The models were acquired with a stereophotogrammetric device and edited to isolate the iliac region. For each class, the left model was mirrored and superimposed on the right one, and the point-to-point distance of each superimposition was recorded in terms of Root-Mean-Square (RMS) value. Seventy-seven superimpositions between true-matches and 30 superimpositions between mismatches (10 for each class of preservation) were performed. Statistical differences between RMS values of the three classes were investigated with a two-way analysis of variance (ANOVA) test (p < 0.05).
ANOVA revealed significant main effects for both “Type of superimposition” (match or mismatch), and “Class of preservation”, both with a p-value < 0.001. Significant differences between matches and mismatches within each level of “Class of preservation” were found (p < 0.001), with matches having significantly lower RMS values compared to mismatches. Overall, the results showed the interaction between the Type of superimposition and Class of preservation, where the preservation state significantly affects the RMS outcome of the superimposition.
For the first time, this test systematically demonstrated that the RMS distance values of bones affected by postmortem damage differ according to the extent of the damage. Medium to considerable damage intensively affect the performance of the 3D-3D pair-matching, leaving this issue unresolved for future studies.