This paper reports on a project that sought to increase the number of joints able to be sorted by deviation analysis, supporting previous research that demonstrates its value as a technique for resolving commingled human remains.
This study follows up on previous research conducted by Litavec on sorting commingled sacroiliac joints using deviation analysis. In the present report, the results of this technique are expanded to separating commingled first tarsometatarsal and atlantoaxial joints. Following the methods of Litavec, virtual models were created at the University of Tennessee-Knoxville Donated Skeletal Collection from 69 atlases and axes, 66 first metatarsals, and 64 medial cuneiforms using an EinScan-Pro 2x+ Handheld Surface Scanner. The shape and congruency of the articular surfaces were analyzed using a deviation analysis in Geomagic Wrap 2021. Receiver operating characteristic curves were conducted on reference samples composed of 200 commingled and non-commingled bone pairs per joint to identify threshold values for sorting the remains. Validation samples of 225 pairs were subsequently analyzed to identify the efficacy of this method on a sample of unknown individuals. The statistical analyses confirmed that deviation analysis values from commingled first tarsometatarsal and atlantoaxial joint pairs were significantly larger than those from non-commingled individuals (p < 0.0001). Eighty-seven percent to 92 percent of first tarsometatarsal pairs and 63 percent to 66 percent of atlantoaxial joint pairs were correctly sorted based on the selected threshold values. This study increases the number of joints able to be sorted by deviation analysis and reemphasizes its value as a technique for resolving commingled human remains. (Published Abstract Provided)