| Background and objective:Traumatic Brain Injury (TBI) is the frequently-occurring disease in the neurosurgery department. The mechanical deformation by the vulnerant force affected on the head induces the primary injury and the secondary injury. The TBI can cause the extensive clinical symptoms and dysfunctions. It has been the social public health problem now, because it is one of the major causes of the human death and disability. The understanding of the biomechanics in the different brain tissue can effectively estimate and prevent the TBI degree. But the detailed machanism of the human TBI is not completely illuminated. The stress-strain relation in the human TBI is difficult to be acquired, and the biomechanics research of brain injury in the living body is impossible. Over the half-century years, many different experimental methods are utilized and developed in the brain injuries such as animal experiment, corpse experiment and biomaterial experiment. Due to the difficulties of direct obtaining and analysis the different structures biomechanics with these relative experimental method, most researchers attempt to utilize the mathematical model for the TBI field. The finite element method (FEM) is the one of numerical method in structural analysis. It has been applied to solve the mechanics response of the several materials with the adventages in the analysis of complicated geometry shape and anisotropic material.For the past few years, with the development of computer technologies and the higher mathematics, especially the computer aided design system (CAD), more and more finite element models are created and used to analyse the biomechanics changes in the human TBI. Some FEMs include the detailed human anatomical structures for the different tissues stress-strain research. But the human brain finite element models have some disadvantages in the TBI simulation and analysis. On one hand there are several limitations in the acquisition of the different human brain biomechanics parameter, on the other hand the analytic datas of the human brain finite element models are difficult to compare with datas of the actual human TBI. So this study uses the rat as the research object to develop its brain three dimensional (3-D) finite element model. With this model the biomechanics relationships between the different tissues are analysed. Also this study comfirms the changes of maximum principal strain (MPS) is related with the rat cortex injury. The further research purpose is to predict the neuron injury in TBI. The research work includes:1. Establishment of the rat controlled cortical injury (CCI) model and the injury analysis of the cortex and the hippocampusObjective:Establishment of the rat controlled cortical injury model and the injury analysis of the cortex and the hippocampusMethods:We establish the rat CCI model using the special CCI device with the stereotaxis equipment. The impact velocity is 4m/s, the impact depth is 2.4mm, 2.8mm and 3.2mm. The animals are randomly devided into control group and groups at 2.4mm,2.8mm and 3.2mm. To evaluate the injury of the cortex and the hippocampus of every experiment groups, Cresyl Violet staining is performed to calculate the cortex injury area and hippocampus survival neurons.Result:The rat CCI models are established successfully. The cortex of impacted side is sharply damaged in the different impacted depth. The defective percent area in cortex of the contralateral hemisphere is 3.65±2.11 in the 2.4mm group,7.83±2.53 in the 2.8mm group and 12.85±3.02 in the 3.2mm group. The surviving neuron number in ipsilateral CA3 hippocampus 35.67±6.38/hpf in controlled group,26.50±4.18/hpf in 2.4mm group,17.67±4.08/hpf in 2.8mm,. In 3.2mm group, the impactor penetrate into the pyramidal layer, we do not count the neurons number.Conclusion:The rat CCI model is the scientific and advisable animalmodel for the TBI research with the controlled impacted force parameter. In the rat CCI model, the severe brain injury parameter should set that the impacted depth is 2.8mm with the 4m/s velocity.2. Construction of the 3-D finite element model of normal rat brainObjective:Construction of the 3-D finite element model of normal rat brainMethods:According to the normal rat stereotactic anatomical atlas, we confirm the reconstruction targets with the 25 coronal slice. Each slice anatomies are processed and digitizing with Photoshop CS software. The detailed anatomies includes the cortex, subcortex, hippocampus and lateral ventricles structures. All digital atlas are located in the stereoscopic coordinate. Each slice is saved as the TIFF format. Then with the SolidWorks 2007 software, two-dimension (2-D) draw of each slice is done according to the anatomical shapes and stereoscopic coordinates of each tissue. Each cross section is connected with the lofting order in SolidWorks 2007 software, the hippocampus and lateral ventricle structures are exsected. The assembly is built with the each part and saved as the xt format. Then this format is imported into the Unigraphics NX 6.0 software. In this software the impactor is constructed and the impactor tips located on the cortex surface. The impactor diameter is 5mm, the impacted direction is perpendicular. Then the total model is imported into the ANSYS Workbench software to define the elements type with solid 164 tetrahedron. The mesh is constructed according to the mechanical environment default setup. The mesh density in the most of ipsilateral part is 0.3mm, in the impacted side hippocampus and contralateral hemisphere is 0.5mm.Result:Construction of the 3-D finite element model of normal rat brain is completed. The model includes the cortex, subcortex, hippocampus and lateral ventricles structures. The FEM has a good geometric similarity with the anatomical atlas. The total number of elements was 205348, the total number of nodes is 40412.Conclusion:The rat anatomical atlas in the stereoscopic coordinate provides a effective, simple, economic and accurate information for the construction of finite element model of the complex shapes and structures. With the image processing software, CAD software and finite element software, it can be effective to construct the mathematical model of the complicated organism.3. The finite element method simulation of the CCI animals model, the biomechanical analysis and the initial research in prediction of the cortical neuron lossObjective:1) Construction of the simple rat cortex impacted model, simulation of the rat CCI model process, collating of the mesh density and decay constant2)The finite element simulattion and analysis of rat CCI model with the FEM, initial prediction of the neurons injury with the parameter MPS Methods:1)According to the cross section of the 3-D FE model, a plane FE model is constructed in the quadrilateral elements with the simple brain structure and the impactor by the Unigraphics NX 6.0 software. In the ANSYS Workbench software, the Explicit dynamic is built. The viscoelastic material properties of the brain grey matter and white matter are set. The properties include density, Young's modulus, poisson ratio, viscosity shear modulus and elasticity shear modulus. Two different areas are marked to compare with the different mesh density. The density includes 2mm?1mm?0.8mm?0.3mm?0.2m mand 0.1mm. According the 0.5mm mesh density, the different decay constants are choosed to analyse the MPS changes. The decay constants include 5ms?10ms?20ms and 40ms.2)The impacted parameters including the material attribute, impactor diameter and the impacted velocity according to the animal experiment are set in the Ansys WorkBench software. The impacted depth is 2.4mm,2.8mm and 3.2mm. The viscoelastic material properties of the brain grey matter and hippocampus and the ventricle are set. The properties include density, Young's modulus, poisson ratio, viscosity shear modulus and elasticity shear modulus. The boundary condition is set according to the skull density. And four regions in cortex, subcortex, hippocampus and ventricle are choosed as the regions of interest. In different impacted depths each regions has the different stress, strain and deformation relationship. The linear regression is carried out between cortex injury in the animal experiment and the MPSs in the FEA predicted for the regions of cortex.Result:1)There is the different MPS in the impacted simulation with different mesh density. The analysis time of the impact in the model increases obviously with the mesh density decreases. Total average MPS increases when the mesh density decreases, and the increase trend grows down. The MPS in selected region increases with the lower mesh density. The difference between 2mm mesh density and 1mm density was 29.5%. The differences between 0.3mm and 0.2mm,0.8mm were 6.5% and 10.4%. The MPSs do not have a large changes with the different decay constants. The MPS in surface selected region is 0.8165?0.8360?0.8423.0.8420 respectively, in deeper part is 0.0852?0.0852?0.0851?0.0850 respectively.2)The rat 3-D FE model simulates the CCI animal experiment process well in the different impacted depth. The highest value of von Mises stress 0.57064Mpa, 0.64684Mpa, 3.0449Mpa were respectively in the different vimpacted depths. When the impactor reachs the maximum depth, the area under the impactor has the largest deformations, and decrease in the radial directions, the deep part structure has the relative small deformations. And the highest value of strain in increased impacted depths were 1.1179,1.2535 and 1.5571. In 3.2mm depth the high peak of strain continued until the computer simulation analysis ended. In the four interest regions, the hightest MPS value occurred on the cortex in impacted side. All highest values in different groups were present within the foreset analysis time. In 2.4mm and 3.2mm depths simulation, the surface cortex was present the higher value than subcortex, it was 1.2%and 11.1%respectively. In the another depth the subcortex was 4.4%higher than surface cortex. The lateral ventricles had the lower change range in MPS. In 2.4mm and 2.8mm depths groups, the hippocampus was taken the higher MPS than lateral ventricles. In the another group the ventricle was taken more higher. The differences were 29.2%,40.8%and 27.8%. The initial research in linear regression between the MPSs predicted by the FE model and the neuron injury of cortex in animals experiment. The correlation coefficient R~2 were 0.9953 and 0.9685.Conclusion:1)The analysis result is more exactly and the analysis time is more longer when the mesh density decreases. The mesh density in 0.3mm is good for the precision and reasonable computing time. The decay constant in 20ms has little influence in our study.2)In the simulation of the CCI animal model, the different brain tissues have the complicated stress, strain and deformation relationship when the cortex has been impacted. The differences in the biomechanics responses induce the different levels of the TBIs in the primary injury and secondly injury. In the initial study the MPS can be the parameter that predicts the neurons injury in TBIs. More animals experiments and more different pre-processing conditions FE models should be developed to implore the feasibility of MPS acted as the quantitate parameter of cortical neuron injury in TBI. |
PDF Full Text Request