The Effect of Variable and Uniform Densities on the Inertial Properties of Cadaver Segments, a Comparison of Frozen and Thawed Densities, and Segment Densities Related to Endomorphy and Ectomorphy

This study was designed to answer three questions. Is there a significant difference between the inertial properties of the segments, such as center of mass location and magnitude of principal moments of inertia, when variable density within the segment is compared to a comparable uniform density model? Is there a significant difference between frozen and thawed segment densities? Is there a significant difference in the density of ectomorphic and endomorphic body types? Two cadavers of different body mass indices were segmented into 16 segments each and further cross-sectioned into horizontal slices 1 to 2 cm. thick. Densities of frozen and thawed segments and sections were measured by immersion, and variable and uniform density models were examined regarding inertial properties. A stacked ellipse model having identical geometries for both conditions was used to attempt to answer these questions. Variable and uniform densities yielded results that differed by a maximum of about 0.3% for all segments and which did not approach significance in either center of mass location or magnitude of principal moments of inertia. It was concluded that models that include variable density (such as DEXA, MRI, CT) increase the expense of estimating inertial properties without any appreciable gain in accuracy over less expensive models based on topography alone. There were significant density differences (p<0.01) when comparing ectomorphy with endomorphy. It is suggested that it may be useful to consider density differences due to body somatotypes when performing motion analysis. Comparison of frozen density and thawed density showed significant differences in both subjects (p<0.01), and applied to all segments. It is also suggested that since frozen density is significantly different from thawed density, that thawed density, being more comparable to analyses in vivo, should always be used in analysis of movement.