- Email: [email protected]
Biomechanics and early pathology of spinal cord injury depend on direction of vertebral fracture dislocation
$150
Journal o f Biomechanics 2006, Vol. 39 (Suppl 1)
the paediatric population; where spinal cord injuries result without radiographic abnormality of the spinal column. 5325 Tu, 08:45-09:00 (P16) Strain energy density variations o f o s t e o p o r o t i c spine column after bone segment augmentation - An in vitro porcine biomechanical model C.-K. Chiang, J.-L. Wang. Institute ef Biomedical Engineering, National Taiwan University, Taipei, Taiwan Introduction: Percutaneous vertebroplasty (PV) is one of the popular surgeries for osteoporotic compression fracture. Long-term clinical observations showed that there are new fractures developed after undergoing PV surgery [1]. We hypothesize new fractures is due to distribution change of strain energy density (SED) after PV. In this study we try to evaluate SED variations of damaged vertebrae and adjacent ones. To our knowledge, there have been no reports to date about using energy methods to explore the load transmission mechanism. Materials and Methods: Twelve fresh porcine spine specimens (L1-L5) were used. The 3-axial strain gages were glued on the three vertebral bodies (L2/L3/L4). We drilled a 15 mm diameter hole on the L3 vertebrae to mimic the strength weakening of osteoporotic and damaged vertebral body. We filled the drilled hole with bone cement to mimic the augmentation of standard PV procedure. A "drop-tower type" testing apparatus was used for the testing. 3.6 J input energy was applied on the specimen. The two principal stresses of vertebral body can be calculated from the measurement of strain gage rosette. The total SED hence can be obtained from the principal stresses and stress invariant. Results: The averaged peak loading is 1570 (SD 90) N. The SEDtotal/N of damaged vertebra (L3) is 50.74 (SD 28.57) j/m3/N and decreases significantly to 9.05 (SD 4.82) j/m3/N after augmentation (p=0.005). The SEDtotal/N of upper adjacent vertebra (L2) increases significantly from 5.82 (SD 1.71) to 7.2 (SD 2.3) j/m3/N (p =0.009). Conclusion: The SED of cemented vertebra is reduced and the input energy of motion segment remains the same. Therefore the adjacent untreated vertebras have to absorb the extra energy and hereafter at higher risk of failure. This mechanism may explain the clinical observation of adjacent vertebral failure after PV. References [1] Uppin AA, Hirsch JA, Centenera LV, et al. Occurrence of new vertebral body fracture after percutaneous vertebroplasty in patients with osteoporosis. Radiology 2003; 226: 119-24.
5468 Tu, 09:00-09:15 (P16) The effect o f cerebrospinal fluid on the biomechanics o f spinal cord: an in vitro animal model study C.F. Jones 1,2, S.G. Reed 2, P.A. Cripton 2, R.M. Hall 1. 1School efMechanical Engineering, University of Leeds, Leeds, United Kingdom, 2Injury Biomechanics Laboratory, School of Mechanical Engineering, University of British Columbia, Vancouver, Canada Spinal cord injury affects more than 25 people/million annually in the United States and has significant economic and social costs [Burke et al 2001]. Current research seeks to characterize the mechanical behaviour of the spinal cord and describe its interactions with spinal column components during the injury event. Knowledge of the column-cord interaction and its relationship to neurological injury could inform preventative strategies and clinical interventions as well as aid the validation of finite element models. The current study extended a previously developed in vitro animal model of the burst fracture process [Hall et al, 2005]. In particular, the study investigated the effect of cerebrospinal fluid (CSF) on the biomechanics of the cord when subjected to high speed transverse impact. The impact of a propelled bone fragment analogue with the animal cord model was recorded with high speed video and the images analysed to determine the deformation trajectory. Each cord was tested with dura and CSF, with dura only and bare. The effect of these states on spatial and temporal descriptors of the deformation trajectory was assessed. The dura was found to have no significant effect on deformation behaviour. Cord deformation was significantly reduced, although not eliminated, in the presence of CSF when compared to the bare state. The time to achieve maximum deformation and the duration of deformation were generally increased in the presence of CSF, though not statistically significantly, which may indicate a reduction in the cord-fragment interaction force for a given impulse. The difficulty in reconciling data to obtain quantitative measures of cord deformation when surrounded by dura and CSF is discussed. This study indicates that while the protective mechanism of CSF may not fully extend to the high energy impact characteristic of a burst fracture, it may contribute to a lessening of cord deformation and applied force.
Oral Presentations References Burke, D.A., et al. (2001). Incidence rates and populations at risk of spinal cord injury: A regional study, Spinal Cord, 39: 274-8. Hall, R.M. et al. (2005). In-Vitro Modelling of the Column Cord Interaction During Spinal Burst Fractures, Proc. of British Orthopaedic Research Society, Stanmore, UK.
6317 Tu, 09:15-09:30 (P16) Development and evaluation o f a continuum neck muscle model S. Hedenstierna 1, P. Halldin 1, K. Brolin 1, H. von Hoist 1,2. 1Department of Neuronics, Royal Institute of Technology, Stockholm, Sweden, 2Division of Neurosurgery, Karolinska Hospital, Stockholm, Sweden Introduction: The biomechanics of the human neck has been studied using numerical models for many years and recently the cervical musculature has gained in interest. Most cervical models of today use spring-elements as muscles. A detailed solid-element muscle model will add properties such as tissue inertia, compressive stiffness, and friction between tissues into the model and should therefore better predict injuries to the cervical tissues. The aim of this study was to determine how a continuum muscle model influences the injury prediction in a human neck FE model compared to a discrete muscle model. Method: The cervical muscles were modeled with finite elements using a detailed 3D anatomical geometry. The geometry was digitized from MR images of 50 th percentile males. The MR images were segmented and interpolated to generate a three-dimensional surface for each muscle. All surfaces were positioned relative to the FE KTH neck model [Brolin et al. 2005] in line with anatomical data from the literature and adjusted to a normal lordosis of a sitting person before meshing. The passive properties were modeled using solid elements and a non-linear, viscoelastic continuum material model. The active properties were modeled separately using discrete elements and a Hilltype material model. The material modeling was validated for a tensile test of a rabbit muscle for strains below 30%. Tissue injury predictions in ligaments, discs, and vertebrae were evaluated and compared to an existing discrete muscle model [Brolin et al. 2005]. Results: The KTH neck model with continuum musculature generates a motion pattern that takes the compressive stiffness, tissue inertia and the material properties of non-linearity and viscosity into account. A continuum muscle model improved the boundary conditions for the vertebral column compared to a discrete model. A continuum model is suggested for simulations of excessive motions and 3D oblique impacts. References Brolin K., Halldin R, Leijonhufvud I. (2005). Traffic Injury Prevention 6(1): 67-76.
6756 Tu, 09:30-09:45 (P16) Biomechanics and early pathology o f spinal cord injury depend on direction o f vertebral fracture dislocation E.C. Clarke 1,2, A.M. Choo 3, J. Liu 3, C.K. Lam 3, L.E. Bilston 2, W. Tetzlaff3, T.R. Oxland 3 . 1University of Sydney, Sydney, Australia, 2prince of Wales Medical Research Institute, Sydney, Australia, 3University of British Columbia, Vancouver, Canada Fracture dislocation is the most common cause of spinal cord injury in human adults. Two experimental models have recently been independently developed to study the biomechanics and neuropathology of SCI following closed column fracture dislocation. The aim of this study was to compare the biomechanics and early neuropathology development of SCI following experimental fracture dislocation of the spinal column in two different directions. This is the first time that the same injury mechanism has been investigated in different loading directions. Thoracolumbar vertebrae T12-L1 were dislocated dorso-ventrally or laterally by 9mm at 220 mm/s in deeply anaesthetised rats. Load and vertebral displacement were recorded. Animals were sacrificed at 1, 3 or 6 hours following injury. Spinal cord sections were stained to detect; haemorrhage, axonal damage and neuronal cell death. Vertebral end-plate fracture was produced in all animals. Fracture load and maximum load were similar for both models, however displacement at failure was significantly higher (p < 0.01 ) for dorso-ventral than lateral dislocation. The amount of tissue damage was greater for dorso-ventral than lateral dislocation. Axonal damage, neuronal cell death and physical damage to the spinal cord were greatest in a narrow band across the spinal cord at an angle of approximately 45 degrees to the direction of loading, in the plane of loading. This may indicate that the damage occurred in the direction of maximum shear strain. This study shows that, for the same mechanism of injury, the pattern of pathology development is similar for both loading directions, however the direction can affect the severity of the spinal cord injury. These results complement literature reports that degree of dislocation and neurological deficit are worse for anterior than lateral dislocation in humans.
Journal o f Biomechanics 2006, Vol. 39 (Suppl 1)
the paediatric population; where spinal cord injuries result without radiographic abnormality of the spinal column. 5325 Tu, 08:45-09:00 (P16) Strain energy density variations o f o s t e o p o r o t i c spine column after bone segment augmentation - An in vitro porcine biomechanical model C.-K. Chiang, J.-L. Wang. Institute ef Biomedical Engineering, National Taiwan University, Taipei, Taiwan Introduction: Percutaneous vertebroplasty (PV) is one of the popular surgeries for osteoporotic compression fracture. Long-term clinical observations showed that there are new fractures developed after undergoing PV surgery [1]. We hypothesize new fractures is due to distribution change of strain energy density (SED) after PV. In this study we try to evaluate SED variations of damaged vertebrae and adjacent ones. To our knowledge, there have been no reports to date about using energy methods to explore the load transmission mechanism. Materials and Methods: Twelve fresh porcine spine specimens (L1-L5) were used. The 3-axial strain gages were glued on the three vertebral bodies (L2/L3/L4). We drilled a 15 mm diameter hole on the L3 vertebrae to mimic the strength weakening of osteoporotic and damaged vertebral body. We filled the drilled hole with bone cement to mimic the augmentation of standard PV procedure. A "drop-tower type" testing apparatus was used for the testing. 3.6 J input energy was applied on the specimen. The two principal stresses of vertebral body can be calculated from the measurement of strain gage rosette. The total SED hence can be obtained from the principal stresses and stress invariant. Results: The averaged peak loading is 1570 (SD 90) N. The SEDtotal/N of damaged vertebra (L3) is 50.74 (SD 28.57) j/m3/N and decreases significantly to 9.05 (SD 4.82) j/m3/N after augmentation (p=0.005). The SEDtotal/N of upper adjacent vertebra (L2) increases significantly from 5.82 (SD 1.71) to 7.2 (SD 2.3) j/m3/N (p =0.009). Conclusion: The SED of cemented vertebra is reduced and the input energy of motion segment remains the same. Therefore the adjacent untreated vertebras have to absorb the extra energy and hereafter at higher risk of failure. This mechanism may explain the clinical observation of adjacent vertebral failure after PV. References [1] Uppin AA, Hirsch JA, Centenera LV, et al. Occurrence of new vertebral body fracture after percutaneous vertebroplasty in patients with osteoporosis. Radiology 2003; 226: 119-24.
5468 Tu, 09:00-09:15 (P16) The effect o f cerebrospinal fluid on the biomechanics o f spinal cord: an in vitro animal model study C.F. Jones 1,2, S.G. Reed 2, P.A. Cripton 2, R.M. Hall 1. 1School efMechanical Engineering, University of Leeds, Leeds, United Kingdom, 2Injury Biomechanics Laboratory, School of Mechanical Engineering, University of British Columbia, Vancouver, Canada Spinal cord injury affects more than 25 people/million annually in the United States and has significant economic and social costs [Burke et al 2001]. Current research seeks to characterize the mechanical behaviour of the spinal cord and describe its interactions with spinal column components during the injury event. Knowledge of the column-cord interaction and its relationship to neurological injury could inform preventative strategies and clinical interventions as well as aid the validation of finite element models. The current study extended a previously developed in vitro animal model of the burst fracture process [Hall et al, 2005]. In particular, the study investigated the effect of cerebrospinal fluid (CSF) on the biomechanics of the cord when subjected to high speed transverse impact. The impact of a propelled bone fragment analogue with the animal cord model was recorded with high speed video and the images analysed to determine the deformation trajectory. Each cord was tested with dura and CSF, with dura only and bare. The effect of these states on spatial and temporal descriptors of the deformation trajectory was assessed. The dura was found to have no significant effect on deformation behaviour. Cord deformation was significantly reduced, although not eliminated, in the presence of CSF when compared to the bare state. The time to achieve maximum deformation and the duration of deformation were generally increased in the presence of CSF, though not statistically significantly, which may indicate a reduction in the cord-fragment interaction force for a given impulse. The difficulty in reconciling data to obtain quantitative measures of cord deformation when surrounded by dura and CSF is discussed. This study indicates that while the protective mechanism of CSF may not fully extend to the high energy impact characteristic of a burst fracture, it may contribute to a lessening of cord deformation and applied force.
Oral Presentations References Burke, D.A., et al. (2001). Incidence rates and populations at risk of spinal cord injury: A regional study, Spinal Cord, 39: 274-8. Hall, R.M. et al. (2005). In-Vitro Modelling of the Column Cord Interaction During Spinal Burst Fractures, Proc. of British Orthopaedic Research Society, Stanmore, UK.
6317 Tu, 09:15-09:30 (P16) Development and evaluation o f a continuum neck muscle model S. Hedenstierna 1, P. Halldin 1, K. Brolin 1, H. von Hoist 1,2. 1Department of Neuronics, Royal Institute of Technology, Stockholm, Sweden, 2Division of Neurosurgery, Karolinska Hospital, Stockholm, Sweden Introduction: The biomechanics of the human neck has been studied using numerical models for many years and recently the cervical musculature has gained in interest. Most cervical models of today use spring-elements as muscles. A detailed solid-element muscle model will add properties such as tissue inertia, compressive stiffness, and friction between tissues into the model and should therefore better predict injuries to the cervical tissues. The aim of this study was to determine how a continuum muscle model influences the injury prediction in a human neck FE model compared to a discrete muscle model. Method: The cervical muscles were modeled with finite elements using a detailed 3D anatomical geometry. The geometry was digitized from MR images of 50 th percentile males. The MR images were segmented and interpolated to generate a three-dimensional surface for each muscle. All surfaces were positioned relative to the FE KTH neck model [Brolin et al. 2005] in line with anatomical data from the literature and adjusted to a normal lordosis of a sitting person before meshing. The passive properties were modeled using solid elements and a non-linear, viscoelastic continuum material model. The active properties were modeled separately using discrete elements and a Hilltype material model. The material modeling was validated for a tensile test of a rabbit muscle for strains below 30%. Tissue injury predictions in ligaments, discs, and vertebrae were evaluated and compared to an existing discrete muscle model [Brolin et al. 2005]. Results: The KTH neck model with continuum musculature generates a motion pattern that takes the compressive stiffness, tissue inertia and the material properties of non-linearity and viscosity into account. A continuum muscle model improved the boundary conditions for the vertebral column compared to a discrete model. A continuum model is suggested for simulations of excessive motions and 3D oblique impacts. References Brolin K., Halldin R, Leijonhufvud I. (2005). Traffic Injury Prevention 6(1): 67-76.
6756 Tu, 09:30-09:45 (P16) Biomechanics and early pathology o f spinal cord injury depend on direction o f vertebral fracture dislocation E.C. Clarke 1,2, A.M. Choo 3, J. Liu 3, C.K. Lam 3, L.E. Bilston 2, W. Tetzlaff3, T.R. Oxland 3 . 1University of Sydney, Sydney, Australia, 2prince of Wales Medical Research Institute, Sydney, Australia, 3University of British Columbia, Vancouver, Canada Fracture dislocation is the most common cause of spinal cord injury in human adults. Two experimental models have recently been independently developed to study the biomechanics and neuropathology of SCI following closed column fracture dislocation. The aim of this study was to compare the biomechanics and early neuropathology development of SCI following experimental fracture dislocation of the spinal column in two different directions. This is the first time that the same injury mechanism has been investigated in different loading directions. Thoracolumbar vertebrae T12-L1 were dislocated dorso-ventrally or laterally by 9mm at 220 mm/s in deeply anaesthetised rats. Load and vertebral displacement were recorded. Animals were sacrificed at 1, 3 or 6 hours following injury. Spinal cord sections were stained to detect; haemorrhage, axonal damage and neuronal cell death. Vertebral end-plate fracture was produced in all animals. Fracture load and maximum load were similar for both models, however displacement at failure was significantly higher (p < 0.01 ) for dorso-ventral than lateral dislocation. The amount of tissue damage was greater for dorso-ventral than lateral dislocation. Axonal damage, neuronal cell death and physical damage to the spinal cord were greatest in a narrow band across the spinal cord at an angle of approximately 45 degrees to the direction of loading, in the plane of loading. This may indicate that the damage occurred in the direction of maximum shear strain. This study shows that, for the same mechanism of injury, the pattern of pathology development is similar for both loading directions, however the direction can affect the severity of the spinal cord injury. These results complement literature reports that degree of dislocation and neurological deficit are worse for anterior than lateral dislocation in humans.