An Imaging-based Disease Model for Understanding Bone Health

点击翻译按钮获取中文摘要

Osteoporosis, defined by a loss of bone mineral density, affects both men and women of all ages, and current projections indicate that the prevalence of this disease will grow dramatically over the next decade given the aging American population. However, it is increasingly recognized that bone density is not the only metric for bone health: other bone quality features, such as trabecular micro-architecture and shape, play a significant role in determining an individual's risk forfracture. Understanding bone health - both density and quality - at the peak of skeletal development can provide a better understanding of the starting point for osteoporosis. This project addresses the development of a disease model for bone development to better comprehend the genesis of osteoporosis. Developments include: 1) the use of quantitative magnetic resonance (qMR) to characterize bone development in a young adult normal population; 2) the further characterization of trabecular bone using qMR to assess bone quality; 3) the development of a (causal) Bayesian belief network (BBN) for the domain of bone development/health, integrating qMR values, imaging, and other observational data into a unified model; and 4) a visualization framework that uses the BBN to guide the display of medical data and explain phenomena, in addition to allowing users to directly query the belief network. Integrated displays using these methodologies are supported, in addition to investigation of graphical techniques for exploration of large causal models. Two patient populations provide the testbed for this effort: 1) a normal volunteer population, ages 20-30 years old, that will have dual-emission x-ray absorptiometry (DXA), peripheral quantitative computed tomography (pQCT), and qMR studies to create gender and race/ethnicity stratified atlases of normal bone density and quality; and 2) patients seen at UCLA Clark Urological Center who suffer from chronic renal disease resulting in osteoporotic conditions (e.g., renal osteodystrophy, disuse osteoporosis), thus providinga comparison against the normal component of this study. A sub-study examines the role of vitamin D on bone development within the abnormal population. Evaluation will be twofold: 1) technical evaluation of the research in a controlled environment; and 2) clinical evaluation to test the efficacy of the causal disease model and end user interfaces. Relation to public health. The thrust of this proposal is to lay a unified foundation for creating bone development model, leading to a better understanding of osteoporosis. Tools to assist in the creation of this model, allowing for explanation and exploration of medical data, will be developed.