Modern Day Osteometrics: Testing the Concordance of Postcranial Measurements Collected Manually vs. Digital Methods

The use of medical imaging (CT/MRI) modalities for the collection of osteometric data has become increasingly common in anatomical research. However, the relative accuracy of skeletal measurements collected using these modalities is poorly established. Thus, the purpose of this study was to test the hypothesis that osteometric data collected manually and digitally are reliably comparable. The software program 3D Slicer was chosen for collection of digital measurements because of its widespread usage in modern morphometric research. Maximum femoral length, maximum humeral length, humeral epicondylar breadth, and anteroposterior femoral head diameter (FHD) were collected on 36 humeri and 23 femora. Each of these measurements was taken three times using different methods: (1) physical anthropometry, (2) digitally directly from raw Computed Tomography (CT) scans, and (3) digitally from full-bone 3D models generated from CT scans. Maximum length and epicondylar breadth measurements were all found to be highly comparable between all three measurement methods (Lin's Concordance coefficients 0.96-0.99). Conversely, FHD measurements collected via the two digital methods were found to be poorly comparable to manual measurements (Lin's CC: 0.0832-0.160) with average differences of approximately 7.0 mm (16%). These results suggest that measurements collected digitally are generally comparable to those collected manually using traditional osteometric tools. Methodological discrepancies in measuring FHD may relate to the difficulty of reliably orienting the proximal femur in 3D digital space. Future research into these discrepancies is necessary as FHD is frequently used in regression formulae for estimating stature, body mass, and metabolic requirements.