fragility fracture are of great importance. However, there is little of the study on non-invasive bone strength assay. Finite element analysis is a computer aided tech- nique which is based on the scanning images and simulated the bone strength. Thus, we use this FEA model to evaluate the change of bone strength under dif- ferent physical situation. Objective: To establish the rat’s tibial three-dimensional (3D) finite element analysis model and to evaluate the change of bone strength under different phys- ical status. Methods: A set of consecutive transactional microCT tomography images of 10- month-old rat’s tibiae was selected. The scanning parameter was 16um, 70Kv. The scanning region started from the lowest point of the diaphysis. A proximal tibia 3D finite element analysis model was established by using software Mimics12.0. Ap- propriate pressure of the tibia and the orientation of load on the trabecular bone were applied to the model by using the software Abaqus6.8, to calculate the stress values of all nodes and units of each model. Moreover, the mechanical property of tibiae was assayed by compression test. Using the similar technique, we also estab- lished the 3D finite element analysis model of osteoporotic bone and compared the differences between the sham group and osteoporotic group. Results: (1) The mechanical property of the rat tibiae evaluated using compres- sion test. The results showed that the average BMD of the sham group is 0.302  0.03g/cm 2 , whereas for the osteoporotic (OP) group the average is 0.250  0.04 g/ cm 2 . The BMD of the OP group was significantly lower than that of the sham group (P ! 0.01). From mechanical test, comparing the maximum load of the two groups, it was found that the maximum load of OP group is 108.85  9.63N, however the sham group had a slightly higher value of 121.07  8.34N. (2) The mechanical property of the rat tibiae evaluated by FEA. The BMD of sham group was 1.132 g/cm 3 and of the OP group was 0.644 g/cm 3 . When sub- jected to compression fracture, the proportion of elements that exceeded yield strain in the OP groups significantly decreased 70%, especially in the region-I. Conclusion: Our analysis showed that the stronger material properties are of better strengths of trabecular bone tissue. The results also indicated the 3D finite element analysis model is sensitive for bone strength, thus maybe a ideal surrogate biomarker. Poster Number 062 Assessment of Bone Architecture/Microarchitecture MULTIPARAMETRIC IMAGING OF BONE ARCHITECTURE: A CADAVERIC STUDY Yanfeng Su, MD, candidate, University of Cincinnati, Cincinnati, OH Kenneth L Weiss, MD, University of Cincinnati, Cincinnati, OH; Noah Allen, MD, Candidate, University of Cincinnati, Cincinnati, OH; Numan Suzanne, MD, Candidate, University of Cincinnati, Cincinnati, OH; Mike Hazenfield, MD, University of Cincinnati, Cincinnati, OH; Ron Huston, PhD, University of Cincinnati, Cincinnati, OH; Paula Chmielewski, PhD, Procter & Gamble Pharmaceuticals, Health Care Research Center, Mason, OH, Cincinnati, OH; Jun Ying, PhD, University of Cincinnati, Cincinnati, OH; Nelson B Watts, MD, University of Cincinnati, Cincinnati, OH; Rhonda Strunk, BS, University of Cincinnati, Cincinnati, OH; Lisa C Lemen, PhD, University of Cincinnati, Cincinnati, OH; Gary Gross, PhD, Procter&Gamble Pharmaceuticals, Health Care Research Center, Mason, OH; Thomas Dufresne, PhD, Procter&Gamble Pharmaceuticals, Health Care Research Center, Mason, OH; Joseph Nurre, PhD, Procter&Gamble Pharmaceuticals, Health Care Research Center, Mason, OH; Bubal Borah, PhD, Procter&Gamble Pharmaceuticals, Health Care Research Center, Mason, OH Background: Bone architecture is a major determinant of its strength. Currently, assessing trabecular architecture & the effectiveness of osteoporosis therapy by il- iac crest biopsies is invasive. Imaging may be useful in such assessment, given that a - correlation between vertebral fractures and strength, and correlations between measures of microarchitectures across different sites in body (vertebrae, distal ra- dii) and vertebral strength are assumed in the osteoporosis literature. This study aims to examine such assumptions. Method: Individual lumbar vertebra and radius of 14 formalin-fixed cadavers were dissected out after QCT scan, for mCT & strength analysis. QCT: Whole body multi detector CTs were performed with multiplanar reconstruction and were analyzed for bone mineral density (BMD), and fracture prevalence estimated as Spinal Deformity Index (sum of Genant scores of T9-L5 vertebrae). mCT: Struc- tural analysis of lumbar vertebrae and distal radius included architecture parame- ters such as BV/TV, Conn.D, SMI, DA, BS/BV, marrow star volume, and trabecular thickness (TbTh), number (TbN) & separation (TbSp). Strength: Each vertebra was placed between parallel plates and slowly compressed at 1/8 inch/ min until the force needed to shorten the specimen decreased dramatically. The compressive force and the resulting specimen deformation were continuously re- corded during. From above compiled and plotted data, the strength for each ver- tebra was calculated and log transformed for statistic analysis. Results: Subjects has median (range) of age 78 (39, 98) years and female:male ratio of 10:4. Mean  Std of vertebral strength (log transformed) was 6.01  .66 for 11 subjects of mild fracture(SDI 7); higher than that of 5.03  .87 for the rest of 3 subjects with worse fracture. (p 5 .05). Vertebral strength and BMD shows + correlation(r 5 .87, p ! .001). Correlations between vertebral strength (log trans- formed) and architectural parameters were summarized in table 1. A multivariate linear model showed the vertebral strength (log transformed) could be predicted by many radius architectural parameters and 90% of variability was explained by those predictors. Discussion: Vertebral strength was found associated to the fracture. Architec- tural parameters at both vertebra and radius were evaluated; some showed strong relationships to the vertebral strength. The finding could be clinically promising as it suggested architectural parameters at radius might be able to predict to vertebral strength and eventually the risk of osteoporotic fracture. Poster Number 063 Assessment of Bone Architecture/Microarchitecture ‘‘The Best Technologist Abstract’’ SPINE OSTEO-ARTHROSIS HAS NO EFFECT ON TBS ASSESSMENT: A SITE MATCHED STUDY WITH BMD Renaud Winzenrieth, PhD, Med-Imaps, Pessac, France Remy Dufour, MD, Sainte-Catherine Institute, Avignon, France; Yana Popova Engineer, Med-Imaps, Pessac, France; Didier Hans, PhD, MBA, CCD, CDT, Department of Bone and Joint Diseases, Lausanne University Hospital, Lausanne, Switzerland In the premenopausal population, spinal BMD predicts the risk of any fracture as well as hip BMD. In later life, however spine measurements are confounded by osteo-arthrosis. The latest guidelines published by the ISCD clearly states in case of osteo-arthrosis, it is accepted to exclude a vertebra when there is more than a 1.0 T-score difference between the vertebra in question and adjacent verte- brae. We know that osteo-arthrosis artificially increases DXA BMD measurement proportionally to its severity. TBS is a grey level texture parameter which charac- terizes micro-architecture status of bone independently of BMD. The objective of this study is to investigate the effect of the osteo-arthrosis on measurement of A-P lumbar spines TBS values using DXA imaging. We present a cross-sectional study on 390 Caucasian subjects. Study group was composed of 141 cases presenting arthrosis (according to ISCD definition) only at L4 vertebra with mean age and BMI of 66.0  8.3 years and 25.2.8  3.5 Kg/m 2 respectively and 249 control sub- jects free of arthrosis with mean age and BMI of 64.1  6.9 years and 24.5.8  3.4 Kg/m 2 respectively. Cases were stratified using severity of arthrosis defined by the differences between L3 and L4 expressed in standard deviation of T-score (severity ranges between 1 to 3.5 T-score). In order to validate control and case groups, a comparison between BMD and TBS data of these groups at L1-L3 was done. In addition, TBS values of control subject were compared with French TBS nor- mative data at L1-L4. BMD and TBS were evaluated at AP Spine (L1-L4) with DXA prodigy (GE-Lunar) and TBS iNsightÒ (Med-Imaps). Absolute differences in percent between case and control groups at L1-L3 for BMD and TBS were 2.4 and 3.3% respectively. These differences were lower than their respective least 130 2010 ISCD/NOF Symposium Abstracts Journal of Clinical Densitometry: Assessment of Skeletal Health Volume 13, 2010