基于CT图像的脊椎骨组织分割

Segmentation of vertebra based on CT image

研究的内容是对CT图像切片进行三维重建并从重建的三维脊柱模型上分离出单个的脊椎椎骨。主要包括:对原始带噪CT图像进行噪声去除,对CT图像进行单一阀值结合形态学的分割,并仅保留脊椎部分图形;对分割后的结果进行三维重建,得到整个脊柱的三维模型;使用空间超邻域算法,将单个的脊椎椎骨从脊柱三维模型上分离出来。实验结果表明分离出来的单个椎骨效果良好,自动化程度较高,对虚拟脊椎矫正手术有一定的指导意义。

The research reconstructs the 3-D spinal image and separats 3-D vertebra image from the reconstructed spinal image. Mainly include CT image denoise,2-D segment of spine with single value and morphologic method,3-D reconstruction of seg- mented spinal image,and separated 3-D vertebra image from reconstructed spinal image by a arithmetic which is called beyond neighbor in space.The separated result shows good effect of this way,and it can be used to do some meaningful things for the virtual surgery of spinal.