基于CT图像的肿瘤放射治疗感兴趣区勾画方法

目的：提出一种新的勾画方法,以克制现有一些商用勾画方法资源占用量大且价格昂贵的缺点,进而快速准确地勾画患者CT图像中的一些感兴趣区,以期对于肿瘤放射治疗给予更大帮助。方法：利用人体CT图像的灰度特征,在部分的组织和感兴趣区域边界上结合窗宽窗位等参数、曲线拟合等方法提取患者CT图像中的感兴趣区边界。结果：可以完成人体外轮廓、骨骼、脊髓腔、肺组织、人体体内气腔等多种组织外边界的自动与半自动勾画。结论：该方法可作为治疗计划系统（TPS）的勾画工具应用,也可在图像分析处理工具中使用。由于人体结构的复杂性,该方法对构成过于复杂的组织难以有效地区分。

Method to Get Regions of Interest in Radiotherapy Based on CT Images

Objective To get the regions of interest(ROIs) from computerized tomography(CT) images rapidly and accurately in radiotherapy because some of the commercial software is complicated and expensive.Methods Some ROIs of patients were gotten based on characters of their CT images.Window width,window level and curve fitting were considered in it.Results Some contours of the human body could be gotten automatically and semi-automatically such as body figures,bones,spinal cord,lungs and some caves.Conclusion This m...