Research Of Biological Realistic Model Reconstruction And Visualization Based On Children’s Cervical Vertebra CT Picture

Neck injury is the main reason that causes the death of children in the traffic accident.Children’s neck ligament muscle don’t grow completely, bones are crunchy and theproportion of their heads are larger, which all can make children’s necks hurt. In order toimitate the biomechanical response of children’s neck in the traffic accident veritably, weshould rebuild Children’s biological biological realistic model of cervical vertebra. Accordingto the analysis of the present reconstructive method, this thesis goes deep into researching thecharacteristics of Children’s CT images of cervical vertebra to realize the way of rebuildingChildren’s biological biological realistic model of cervical vertebraIn the process of the reconstruction of Children’s biological biological realistic model ofcervical vertebra, there are two main problems. One is how to segment the cervical regionexactly from the children’s CT images with fuzzy boundary. The other is how to make3dreconstruction for the extracted cervical region.At first, as for the problem of the segmentation of children’s CT images. Because ofChildren’s CT images with the characteristics of fuzzy boundary, unobvious contrast, andquite gentle gradient and so on, before segmenting the images, we should make imagespretreatment. This thesis uses the median filter and window technology to make filter andenhancement processing for the images. Then it makes segmentation processing by using thecommon CT images segmentation method after pretreatment. We can analyze and know that itis effective to segment Children’s CT images of cervical vertebra by using Fuzzy c-meansalgorithm, but this algorithm is sensitive to initial clustering center. In order to solve thisproblem, this article advocates CT images segmentation algorithm by combining Ant ColonyClustering Algorithm which has the characteristics of parallelism and global searchingcapability with Fuzzy c-means algorithm. Using this algorithm to segment children’s CTimages can solve the defect of Fuzzy c-means algorithm, and can reduce the times of imagesegmentation iteration. It improves the efficiency of the segmentation effectivelySecondly, it makes3d reconstruction for the extracted cervical region. Image reconstruction includes Surface reconstruction and body reconstruction. This thesis mainlyresearches marching cube method in the process of surface rendering, puts forward theimprovement way based on contour element construction and rebuild Children’s biologicallifelike model of cervical vertebra by using the Marching Cube algorithm and Ray Castingalgorithm.After solving the both problems above, it develops the Biological lifelike modelreconstruction system based on Children CT images of cervical vertebra by usingVisualization Tool Kit（VTK） on the platform of Visual Studio2008. Not only this systemrealizes the total process from the pretreatment of Children CT images of cervical vertebra to3d reconstruction, but also it is with modularization, integration, and extendibility. So it hasreference value for researching the reconstruction like the model of CT images and it providesfoundation and platform for improving the Biological fidelity of Reconstruction modelfurther.