- Email: [email protected]
Non-invasive determination of long bone structural properties
Abstracts
615
A 3-D,
NONLINEAR, FINITE ELEMENT ANALYSIS OF THE BMBALtiBD HUMAN TIBIA M. Cooper, J. Wasserman Phd, R. Krieg Phd, J. Snider Phd. Departments of Engineering Science and Industrial Engineering, The University of Tennessee, Knoxville, TN 37996, USA. A relatively new topic in the biomedical field is the area of impact biomechanics, which involves the study of trauma to the human body due to impact conditions. The research in this area is primarily experimental, and little finite element analysis has been performed. The research presented in this paper is to apply the finite element method to accurately model a 7.47 m./sec. impact at the tibia1 midshaft in the anterior to posterior direction. It is through the finite element modeling that a better understanding of impact trauma to the huma; lower leg can be attained. A three-dimensional, nonlinear, transient, dynamic, solid mechanics, finite element code is used to analyze the impact condition. The model will investigate fracture propagation and the incurred fracture force in the tibia. In an attempt to verify the finite element analysis , experimental testing matching the computational model conditions was performed on eleven dry, embalmed, human tibias. Fracture force data and fracture propagation trends were investigated in these tests. Results show that the finite element model agrees with the general trends shown experimentally for both fracture force and fracture propagation in the tibia. However, comparison between the experimental and computational models were not entirely possible, and only general trends could be noticed. With the development of an accurate constitutive model of the tibia and better experimental verification, the finite element method can prove to be a valuable tool in injury prediction and the design of injury mitigating devices.
DETERMINATION OF THE ACTUAL ULTRASONIC PATHWAY AND THE WAVELENGTH DEPENDENCE IN RELATION TO THE BONE SPECIMEN AND MICRCXTRUC!‘KJRAL DIMENSIONS Jae-Young Rho* and Kenneth P. Wagner5 *Department of Orthopaedic Surgery and 4Mcchanical Engineering State University of New York at Buffalo, NY 14214, USA The pulse trasmission ultrasonic technique was undertaken to characterize the actual pathway and the wavelength dependence in relation to the bone specimen and microstructural dimensions. Transverse cortical bone velocity can be determined rather than the apparent transverse velocity by using the second order polynominal equation which allows the estimation of the path length as follows: y=l.O02+1_236e-2*x+0.226*x2, where y-Tathlength, x--ratio of medulla to cortex diameter. The average velocity through a single trabecular bone was 2893.8 m/s (SD 152.1), while the mean velocity through cylindrical cancellous specimens was 2704.7 m/s (SD 142.3). Thus, the velocity through the cylindrical specimens was underestimated as 6.5 percent much as that through a single trabecular bone. Inhomogeneity and the variation in anisotropy of cancellous bone may contribute to velocity measuiement error. However, the actual pathway of cancellous bone measurement was slightly longer than the length of the specimens (6.5%). It can be easily postulated that ultrasonic energy has propagated through the smallest passway. The bar and bulk velocities of aluminum calculated from tabulated elastic coeffkients were found to be 5083m/s and 632Omls respectively. The bulk velocities in this expreiment were found to be 6100 m/s. However, a transition between the bulk and bar wave velocity was not exhibited. The results of this study substantiated that dimensions and wavelength should be carefully considered so that the specific velocities and the associated elastic constants can be calculated.
NON-lNVASJVE DElEFMlNATl0N ff LONGBONESTRUCTURALPROPERTIES.
Tammy M. Cleek and RobertT. Whalen NASAlAmes Research Center, Moffett Field, CA 94035. Cross-sectional asymmetry and cuwature found in long bones are related to musculoskeletal force8 generated during daily activity. We have developed a method that can be applii to produce non-uniform structuralmodels of long bones with dual energy x-ray absorptiometty(DXA) data. Principal 8econd moments of area and principal axes of mineralized tissue at each scan cru8s-sectic~1along the length of a bone can be determined by combining attenuation data from 3 non-coplanarDXA beans. Whole bone models are genera&l by combining the section properties from each cross-section. We have investigated this technique to determine sensitivity to (1) scan resolution, (2) cross-sectional shape, (3) orientation to the scan plane, (4) included angle of non-coplanar scans, and (5) isotropy index, Imin/lmax. Axi- and non-axisymmetric aluminum phantoms of known cross-sectional properties and geometries covering normal ranges for long bones were de8igned to validate our approach. phantom designs includedtubes with a tapering outer &meter, an elliptical cross-section, an off-center circular cross-section, and a tapering double helical Qroove. Preliminarywork on a cadaver femur wa8 al8o done. Normal 8pine 8can8 were a8 accurate a8 soane with twice the resolution in obfaining section propertiee. Calculated principal 8eoond moment8 of area were wlthln 34% and principal angles were within 2 degree8 of the expected values for all phantom8 scanned with included angles of 30, 60, and 90 degrees in normal spine mode. Significant changes in principal moment8 and orientationsalong the length of a cadaver femur were observed, althou* the femur was nearly i8otropicin the middiaphysis with respect to sectional properties. The accuracy and insenrllfvfty 04 the algorithm to cmsssectional shape and changing isotmpy values (0.5 to 1.0) when varfous included angle8 were u8ed make thii technique viable for future in viva studies. We befii thi8 method will provide a useful link between ohanges in bone mineral distributionwith changes in daily activity patterns, space fliit, age, and spinal cord injury.
615
A 3-D,
NONLINEAR, FINITE ELEMENT ANALYSIS OF THE BMBALtiBD HUMAN TIBIA M. Cooper, J. Wasserman Phd, R. Krieg Phd, J. Snider Phd. Departments of Engineering Science and Industrial Engineering, The University of Tennessee, Knoxville, TN 37996, USA. A relatively new topic in the biomedical field is the area of impact biomechanics, which involves the study of trauma to the human body due to impact conditions. The research in this area is primarily experimental, and little finite element analysis has been performed. The research presented in this paper is to apply the finite element method to accurately model a 7.47 m./sec. impact at the tibia1 midshaft in the anterior to posterior direction. It is through the finite element modeling that a better understanding of impact trauma to the huma; lower leg can be attained. A three-dimensional, nonlinear, transient, dynamic, solid mechanics, finite element code is used to analyze the impact condition. The model will investigate fracture propagation and the incurred fracture force in the tibia. In an attempt to verify the finite element analysis , experimental testing matching the computational model conditions was performed on eleven dry, embalmed, human tibias. Fracture force data and fracture propagation trends were investigated in these tests. Results show that the finite element model agrees with the general trends shown experimentally for both fracture force and fracture propagation in the tibia. However, comparison between the experimental and computational models were not entirely possible, and only general trends could be noticed. With the development of an accurate constitutive model of the tibia and better experimental verification, the finite element method can prove to be a valuable tool in injury prediction and the design of injury mitigating devices.
DETERMINATION OF THE ACTUAL ULTRASONIC PATHWAY AND THE WAVELENGTH DEPENDENCE IN RELATION TO THE BONE SPECIMEN AND MICRCXTRUC!‘KJRAL DIMENSIONS Jae-Young Rho* and Kenneth P. Wagner5 *Department of Orthopaedic Surgery and 4Mcchanical Engineering State University of New York at Buffalo, NY 14214, USA The pulse trasmission ultrasonic technique was undertaken to characterize the actual pathway and the wavelength dependence in relation to the bone specimen and microstructural dimensions. Transverse cortical bone velocity can be determined rather than the apparent transverse velocity by using the second order polynominal equation which allows the estimation of the path length as follows: y=l.O02+1_236e-2*x+0.226*x2, where y-Tathlength, x--ratio of medulla to cortex diameter. The average velocity through a single trabecular bone was 2893.8 m/s (SD 152.1), while the mean velocity through cylindrical cancellous specimens was 2704.7 m/s (SD 142.3). Thus, the velocity through the cylindrical specimens was underestimated as 6.5 percent much as that through a single trabecular bone. Inhomogeneity and the variation in anisotropy of cancellous bone may contribute to velocity measuiement error. However, the actual pathway of cancellous bone measurement was slightly longer than the length of the specimens (6.5%). It can be easily postulated that ultrasonic energy has propagated through the smallest passway. The bar and bulk velocities of aluminum calculated from tabulated elastic coeffkients were found to be 5083m/s and 632Omls respectively. The bulk velocities in this expreiment were found to be 6100 m/s. However, a transition between the bulk and bar wave velocity was not exhibited. The results of this study substantiated that dimensions and wavelength should be carefully considered so that the specific velocities and the associated elastic constants can be calculated.
NON-lNVASJVE DElEFMlNATl0N ff LONGBONESTRUCTURALPROPERTIES.
Tammy M. Cleek and RobertT. Whalen NASAlAmes Research Center, Moffett Field, CA 94035. Cross-sectional asymmetry and cuwature found in long bones are related to musculoskeletal force8 generated during daily activity. We have developed a method that can be applii to produce non-uniform structuralmodels of long bones with dual energy x-ray absorptiometty(DXA) data. Principal 8econd moments of area and principal axes of mineralized tissue at each scan cru8s-sectic~1along the length of a bone can be determined by combining attenuation data from 3 non-coplanarDXA beans. Whole bone models are genera&l by combining the section properties from each cross-section. We have investigated this technique to determine sensitivity to (1) scan resolution, (2) cross-sectional shape, (3) orientation to the scan plane, (4) included angle of non-coplanar scans, and (5) isotropy index, Imin/lmax. Axi- and non-axisymmetric aluminum phantoms of known cross-sectional properties and geometries covering normal ranges for long bones were de8igned to validate our approach. phantom designs includedtubes with a tapering outer &meter, an elliptical cross-section, an off-center circular cross-section, and a tapering double helical Qroove. Preliminarywork on a cadaver femur wa8 al8o done. Normal 8pine 8can8 were a8 accurate a8 soane with twice the resolution in obfaining section propertiee. Calculated principal 8eoond moment8 of area were wlthln 34% and principal angles were within 2 degree8 of the expected values for all phantom8 scanned with included angles of 30, 60, and 90 degrees in normal spine mode. Significant changes in principal moment8 and orientationsalong the length of a cadaver femur were observed, althou* the femur was nearly i8otropicin the middiaphysis with respect to sectional properties. The accuracy and insenrllfvfty 04 the algorithm to cmsssectional shape and changing isotmpy values (0.5 to 1.0) when varfous included angle8 were u8ed make thii technique viable for future in viva studies. We befii thi8 method will provide a useful link between ohanges in bone mineral distributionwith changes in daily activity patterns, space fliit, age, and spinal cord injury.