A 3D radiological study of age-related quantitative and morphological differences in the human femur: clinical and anthropological applications

© 2010 Dr. Jacqueline Lee Hislop-JambrichThe establishment of a valid scientific instrument for the non-destructive analysis of age-related change in the human femur is essential for the disciplines of clinical medicine and anthropology. The overall aim of this thesis was therefore to develop novel and scientifically verifiable methods for using multiple detector-row computed tomographic (MDCT) data to produce a population-based, morphometric and densitometric record of age-related differences in human femoral bone. The resulting study of contemporary urban Australians is the first to use MDCT data to produce a contemporary model of age-related differences of the human femur. Four independent empirical studies were performed, each of which established the validity of the mortuary-sourced MDCT data, and then reported the age-related radiological findings. The adult population used for this work was taken from the high-resolution MDCT scans performed routinely at the Victorian Institute of Forensic Medicine (VIFM) in Melbourne Victoria Australia. The scans were therefore from those who died unexpectedly in the state of Victoria, and who had no disease or pathology involving the femur. The first study engaged clinical quantitative computed tomography (QCT) software to create a model of the bilateral three-dimensional (3D) densitometric characteristics of the population. The hypothesis was that clinical software may be used effectively on mortuary-based MDCT data to derive population-based assessments of quantitative bone characteristics. Aspects of reproducibility, quality assurance and calibration were all extensively investigated. Clinical ionising-radiation dose estimation software was also used to compare the effective dose delivered during the mortuary MDCT acquisition, to what may be expected in implementing a similar protocol on the living. Results established that a non-destructive QCT technique may be effectively employed using mortuary MDCT data for the purposes of an adult population-based analysis. Results also demonstrated significant differences in gender and age-based expressions of bone mineral and spatial QCT characteristics. Ionising radiation dose calculations showed little variance between the VIFM and theorised clinical protocols. The second study used those same MDCT data to produce 3D image-renders of the bilateral femora. The hypothesis was that there were predictable morphologic age, gender and symmetry related differences within a contemporary predominantly urban adult population spanning 80 years. A series of novel and adapted caliper-type measurements, designed to evaluate age-related differences were implemented. Results were then compared in terms of the physical features of the individuals, such as weight, height and age. A study of femoral symmetry was also completed. Results demonstrated large variability of symmetry in individuals and of deep and surface morphometry across the group generally. Results also demonstrated that irrespective of age, all morphological variables tested were discriminating for gender. The third study used results from an analysis of 17 right-sided femoral specimens from the Melbourne Femur Collection (MFC), to assess the impact of three different MDCT scanning conditions on spatial and quantitative measurements of bone. The hypothesis was that alterations in the scanning conditions and environment would produce little change in the spatial assessments, and consistent variation for the quantitative evaluations. Results demonstrated a strong and predictable correlation in all scanning conditions for both spatial and quantitative assessments. The fourth study produced an epidemiological review of all hip fractures in bodies admitted to the VIFM for autopsy in 2008, as well as a QCT and morphometric assessment from 14 unilateral hip fracture cases. The hypothesis was that the fracture cases would present with dissimilar QCT and morphometric assessments as the normal or un-fractured control group from the first two research chapters. While results demonstrated no difference in morphometric features between the fractured and normal groups, there were significant differences (p