Research On 3D Visualization Of Medical CT Image

Aiming at making medical CT,MRI and other tomographic images more intuitively present and accurately locate the lesion tissue,we take CT tomographic images as the object to study the realization of its 3D visualization and related technologies.In this paper,a series of medical CT images are reconstructed by marching cubes algorithm and ray-casting algorithm respectively,and also improved algorithms are proposed.In the pre-processing stage,the linear interpolation method is improved to calculate the difference between the front and back,left and right,and up and down of the current pixels respectively.The mean of two pixels with minimum difference among these pixel pairs is selected as the gray value of the current pixels.The edge of the image interpolated by this interpolation method is clearer than before.At the same time,a threshold-based segmentation method(such as skull)is adopted when the gray value of the interested part of the image is clearly demarcated from that of other parts;while in the case of the gray value of the image is not obvious,an improved morphological image segmentation method is proposed,which combines mathematical morphology with set theory,and can extract organs or tissues(such as trachea and bronchus)within the human body according to the requirements of the user.In the reconstruction phase,setting different thresholds in marching cubes algorithm and different opacity in ray-casting algorithm will lead to different reconstruction results.In addition,the true 3D display system based on multi-projection rotating body scanning technology is studied in this paper.According to the shape of rotating screen,a method of three-dimensional plane cutting based on medical model is proposed,and the display of medical model in real three-dimensional is simulated.