三维CT图象自动分割法

Automatic Segmentation of Three Dimensional Computerized Tomograph

计算机断层扫描图象（ＣＴ）和核磁共振图象（ＭＲＩ）是诊断疾病和制定治疗方案与手术方案的主要依据之一，而分割ＣＴ和ＭＲＩ图象是定量分析其平面与三维信息的关键之一。本文先采用三维自适应滤波，在保留图象细节的前提下滤除高斯分布噪声和脉冲噪声，然后根据ＣＴ或ＭＲＩ各断层多阈值二值化图象采用区域自动增长算法实现了三维ＣＴ图象和ＭＲＩ图象的自动分割。最后给出了腹部ＣＴ图象的分割结果，该方法分割精度高，速度快。

Computerized Tomograph(CT) and Magnetic Resonance Image(MRI) are an important basis for diagnosis,treatment and surgical operation,and the segmentation of them is one of the key to the quantitative analysis and display of two and three dimensional geometric information.A method of automatic segmentation of three dimensional CT and MRI is introduced in this paper.In the method a three dimensional adaptive filter is used for filtering off Gaussian and impulsive noises and for retaining the details of original image.On the basis of multi level cross sectional segmentation, the automatic three dimensional segmentation is rea lized by using volume growing algorithm. At the end of the paper,the three dimensional segmental results for abdominal CT are given.This method is of high speed and high precision in segmentation and can be widely used in clinical practice.